Vous êtes sur la page 1sur 61

Conceptual Analyses from a

Grothendieckian Perspective
Reflections on
Synthetic Philosophy of Contemporary
Mathematics
by
Fernando Zalamea
Translated from Spanish (2009) by Z. L. Fraser,
Falmouth and New York: Urbanomic/Sequence Press, 2012.

Giuseppe Longo
Centre Cavaills, Rpublique des Savoirs,
CNRS, Collge de France et Ecole Normale Suprieure, Paris,
and Department of Integrative Physiology and Pathobiology,
Tufts University School of Medicine, Boston.1
http://www.di.ens.fr/users/longo

Translated by Fabio Gironi2

1 This text was written during the authors visit in Nantes, hosted by the
stimulating interdisciplinary environment of the Institut dEtudes
Avances (http://www.iea-nantes.fr/), and it is dedicated to Alexander
Grothendieck, whose death occurred during the preparation of its last
draft.
2 The translator would like to thank Robin Mackay for his precious and
generous help in revising this translation.

207

Speculations VI
Introduction

alameas book is as original


as it is belated. It is indeed surprising, if we give it a moments
thought, just how greatly behind schedule philosophical
reflection on contemporary mathematics lags, especially
considering the momentous changes that took place in the
second half of the twentieth century. Zalamea compares this
situation with that of the philosophy of physics: he mentions
DEspagnats work on quantum mechanics, but we could add
several others who, in the last few decades, have elaborated an
extremely timely philosophy of contemporary physics (see
for example Bitbol 2000; Bitbol et al. 2009). As was the case
in biology, philosophy since Kants crucial observations in
the Critique of Judgment, at least has often run ahead of life
sciences, exploring and opening up a space for reflections
that are not derived from or integrated with its contemporary
scientific practice. Some of these reflections are still very much
auspicious today. And indeed, some philosophers today are
saying something truly new about biology.
Often Zalamea points the finger at the hegemony of analytic
philosophy and the associated linguistic turn and the
associated foundationalist projects in mathematics, highlighting the limits of a thought that, by and large, remains
stuck to Hilberts program (1900-1920) and Gdels theorem
(1931) respectively an extremely important program and
an equally important (negative) result, certainly. However, we
should do well to consider that something important happened in the decades that followed, both in mathematics and
in the correlations between the foundations of mathematics
and physics, topics to which Zalamea dedicates several pages
of his book. The conceptual and technical frames invented
by Grothendieck are a fundamental part of these novelties.
At this juncture, I would like to introduce a first personal
consideration: for far too long philosophical reflection on

208

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
mathematics has, with only rare exceptions,3 remained within
the limits of the debate going From Frege to Gdel (as per
the title of a classic collection) a debate at best reaching the
statement of Gdels theorem, or indeed a simplified reduction of it which deprives it of its meaning. The meaning of a
theorem is also (but not only) to be found in its proof, but in
the case of Gdels, it is found only by looking closely to its
proof (see Longo 2010). Thus, with a limited range of references going from Euclid to, at best, the statement of Gdels
theorem, passing through Frege and Hilbert (often skimming over a great deal Riemann and Poincar being cases
in point), for far too long we have debated ontologies and
formalisms, thus moving, as Enriques had already foreseen in
1935, between the Scylla of ontologism and the Charybdis of
formalism, a kind of new scholasticism.4 I think, for example,
that even within Logic, the beautiful results of Normalization
in Impredicative Type Theory (see Girard, 1971, Girard et al.
1989), and of concrete Arithmetical incompleteness, as in the
Kruskal-Friedman Theorem (see Harrington and Simpson
1985) which allow for a breakout from this scholasticism
(see Longo 2011) or indeed the more recent progress in
Set Theory, have not yet received a sufficient and properly
philosophical attention.
Zalameas book is thematically vast. It is truly astounding
to behold the rich range of mathematical themes that are
touched upon, arguably including all of the most important
objects of contemporary exploration. I can only single out a
few of them, in an attempt to hint here to an epistemology
of new interfaces, and to emphasize, for my own account,
3 Among these exceptions, an excellent collection is Mancosu 2008.
4 If we refuse to look for the object of logic in the operations of thought
we open the door to this ontology which scientific philosophy must
to fight as the greatest nonsense. On the other hand, guarding oneself
from the Scylla of ontologism, one falls into the Charybdis of nominalism: how could an empty and tautological system of signs satisfy our
scientific reason? On both sides I see emerging the spectre of a new
scholastics. F. Enriques, Philosophie Scientifique, Actes du Congrs
International de Philosophie Scientifique, Paris, 1935, vol. I-VII.

209

Speculations VI
the timeliness and epistemological relevance of the triadic
relation mathematics-physics-biology which, obviously, is
not the theme of this book.
Modes of Conceptualization, Categories, and Worldviews
6.5.1 Nowadays we may want to overturn Galileos phrase: Is the book
of mathematics written in a natural language?
(Lochak 2015)

I would like to begin with what Zalamea considers, if I am


not misreading his argument, the highest and most revolutionary point reached by post-World War II mathematics:
Grothendiecks work. With a daring table (43) as daring as it
is arbitrary, like any such schematization Zalamea sums up
the principal modes of conceptualization and construction
pertaining to contemporary mathematics []: arithmetical
mixing, geometrization, schematization, structural fluxion
and reflexitity. In his text, he gradually develops the meaning of each of these modes, attributing to Grothendieck alone
the distinction of having contributed to every one of these
forms of mathematical construction.
Before delving deeper into the arguments, and maintaining
a rather survey-like approach (an inevitability when trying to
sum up a book this rich) I think that I can single out the core
node of Zalameas thought in this statement: contemporary
mathematics systematically studies deformations of the representations of concepts (172). In more classical fashion, I
would rephrase this by saying that mathematics is, in primis,
the analysis of invariants and of the transformations that
preserve them (including the analysis of non-preservations,
deformations and symmetry breakings). This does not aim to
be an exhaustive framing of mathematical construction, but
rather the proposal of a different point of view, in opposition
to, for example, the set-theoretical analytical one.
I will also try to show how Grothendieck, in particular, went
beyond this vision of mathematics inherited from Kleins
Erlangen Program and developed by many others (that of
210

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
symmetries, invariants, and transformations). Grothendieck
proposed notions and structures of an intrinsic mathematical purity, free from any contingency requiring proof of
invariance, presented in an highly abstract (yet not formal)
mode, always rich of mathematical sense, particularly thanks
to the analysis of relations with other structures.
Symmetries have clearly laid at the heart of mathematics
since well before Kleins work or before 1931. Indeed we can
trace its centrality to Euclid, whose geometry is entirely constructed out of rotations and translations (symmetry groups as
invariants and as transformations), through Erlangen Program,
Noethers Theorems (1918) and Weyls work between the two
World Wars. I would like to highlight, more than Zalameas
text does, the correlations with the foundations of physics
which these last two mathematicians put at the very core of
their work and, in Weyls case, of his philosophical thought
(see Weyl 1932; 1949; 1952; 1987).
Weyls work profoundly marked the period examined by
Zalamea, moving within a framework which we could legitimately define as that of Category Theory, with frequent
mention, for example, of Topos Theory. Mac Lane, one of
the founders, along with Eilenberg, of this theory, had spent
a year in Gttingen in the early 1930s, in close contact with
Weyl, the great geometer (and mathematician, and physicist). Category Theory, considering the role it plays in the
analysis of invariants and their transformations, is indeed
a profoundly geometrical theory, so much so that it led, in
Grothendieck and Lawvere, to the geometrization of logic,
a topic I shall consider later (see Johnstone 1982; Mac Lane
and Moerdijk 1992). I should also mention (again echoing
Zalamea but with an even stronger emphasis) the role of
physical theory in mathematical invention, with particular
reference to Connes. But we cannot do everything, and I not
being a geometer, and thus unable to adjudicate on many of
Zalameas conceptual and technical analyses shall attempt
to read the text though my contemporary lens, shaped by
several years of cooperation with physicists and biologists
on the interface between the foundations of these disciplines
211

Speculations VI
(see Bailly and Longo 2011; Longo and Montvil 2014).
I am no geometer and Zalameas text, one could say, is dominated by geometrical work, if intended in an extremely broad
and modern sense. It is partially this central role assigned to
geometry that motivates Zalameas vigorous polemic against
analytic philosophy. The latter has done nothing but increase
its focus on linguistic play and logico-formal axiomatics,
without any programmatic relationship with space and the
constructions of physics; without paying attention to the
constitution of mathematics in the world, and to the interface between ourselves and the world described by physics.
Frege and Hilbert, in different ways, both programmatically
wanted to avoid founding mathematics in relation to the
delirium (Frege 1884) or to the challenges of meaning of
non-euclidean geometry and physical (lived and intuited)
space (Hilbert 1901). And they did so for very good reasons.
In order to give certainty to mathematics, it was necessary
to keep in check
1. The dramatic break between the common-sense
intuition of space and a physics in which all that happens are continuous changes in the curvature of space
(Clifford, referring to Reimann 1854).
2. The unpredictability of dynamical systems (Poincar 1892): a result of undecidability of future state of
affairs for non-linear deterministic systems that is, for
formalizable systems of equations at the interface between mathematics and physics (see Longo 2010). It was
considered necessary to make sure that, at least in pure
mathematics, every well-formalized statement could be
decided (Hilbert). This is by principle far, therefore, from
the undecidability and chaos that systems of non-linear
equations had already started to reveal in the context of
physical dynamics.
3. The new and bewildering role played by measurement
in physics, where (classical) approximation or (quantum)
212

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
non-commutativity had introduced unpredictability
(Poincar) and indetermination (Plank) in the interface
between physics and mathematics.
The exactitude of the whole number, a logical and absolute
concept (Frege) and its theory Arithmetic were supposed
to guarantee unshakable certainties (Hilbert), thanks to the
demonstrable coherence and to the formal decidability of
pure mathematics: a far cry from the protean, approximate,
unpredictable, and indeterminate world of physics. And so it
happened that a century of debates on foundations remained
trapped (and for good reasons) between programmatically
meaningless formalisms and Platonist ontologies attempting to deliver a meaning from outside the world; outside,
that is, of the difficult analysis of conceptual construction,
the latter being the real bearer of meaning. It is precisely
this latter kind of project that lies at the heart of Zalameas
philosophical work.
From physics, Zalamea borrows a methodological question:
the great paradigm of Grothendiecks work, with its profound
conception of a relative mathematics [140-141] interspersed
with changes of base of every sort in very general topoi [141
-142], should be fully understood as an Einsteinian turn in
mathematics (270). And so Einsteins Invariantentheorie (as
he preferred to call it) thoroughly becomes part of the method
of this analysis of mathematical construction, broadly based
on invariants and the transformations that preserve them.
It is clear then why this approach assigns a central role to
the notion of the Category. This is not a Newtonian universe
anymore, a unique and absolute framework, the Universe of
Sets, with an absolute origin of time and space (the empty
set). It is rather the realm of a plurality of Categories and of
an analysis of transformations, functors, and natural transformations that allow their correlation (preserving what is
interesting to preserve). Among them, the Category of Sets is
surely one of the most interesting, but just one of many. We
are presented with an open universe of categories, then, to
which new categories are constantly added; new invariants,
213

Speculations VI
and new transformations. Concepts are created by being correlating with existent ones, and by deforming one into the
other, thus enriching them, paying attention to the meaning
(the mathematical meaning, at least) of what is being done.
Thus Zalamea also retrieves an operational relation with
the supposed delirium or disorder we referred to at the interface of geometry with physics : Advanced mathematics
are, by contrast [to the elementary mathematics analyzed in
most philosophical reflections], essentially dynamic, open,
unstable, chaotic [] the geometry of mathematical creativity is replete with unpredictable singularities and vortices
(39). Yet there is an order, a dynamical organization to all
this since, as Lautman puts it, we continuously reconstruct
a hierarchization of mathematical geneses [] a structural
explanation of mathematics applicability to the sensible
universe (58). And this, in particular, is possible thanks to
structural dualities at the heart of any attempt to organize the
world, like those between local/global, whole/part, extrinsic/
intrinsic, continuous/discrete, etc., as Zalamea, writes, again
quoting Lautman (64). Indeed, Lautman intuits a mathematics of structural relations beyond a mathematics of objects
which is to say, he prefigures the path of category theory
(68), which was indeed born just a few years after his death.
The conceptual node that must be added to the analysis
of proof, which was the dominant preoccupation of foundational projects in twentieth-century mathematics, is that of
the analysis of the constitution of concepts and structures
(where these latter are seen as an additional organization
of mathematical concepts).5 This is what Zalamea aims at:
5 Proof theory is an extremely important and elegant branch of mathematics (and by working with its varieties (with and without Types),
its categorial semantics and its applications I have managed to earn
a living for most of my life). However, in philosophy, to omit this or
that pillar of foundational analysis is a typically analytic limit. Corfield
(2003) and Mancosu (2008) have worked to overcome this limit and to
avoid both the Scylla and the Charybdis I mentioned above, by referring
to Mathematical Practice (or Real Mathematics), as if there were a
mathematics which is not a very real praxis: a way to underline the delay
of philosophical reflection on contemporary mathematics, something

214

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
for him, Lautman and Cavaills are frequent points of reference, two philosophers utterly forgotten by logico-linguistic
approaches to mathematics (yet enjoying a more flattering
oblivion than Poincar and Weyl, who have been subject to
offensive caricature as, for example, half-hearted Brouwers
or semi-intuitionists).
I omit several passages and citations from the opening
chapters of the book, where I find myself somewhat perplexed
by what seems to me the excessive space dedicated to those,
like Badiou and Maddy, who place the category of Sets in the
usual role of absolute, Newtonian universe albeit (in Badious
case) with some dynamical inflection. Badiou, for example,
in a recent seminar at the cole Normale Suprieure (Paris)
has explained referring uniquely to the (original) statement
of the Yoneda Lemma that every (locally small) category
is reducible to (embeddable in) the Universe of Sets (Set),
modulo a Topos of prescheaves (on Set). This would definitely
prove the absolute role of Set for mathematics. Now, the proof
of the Lemma yields a more general result. The functional
embedding just described is possible within every Topos
considered as a Universe in which one sees the given (locally
small) category as an object: the embedding is then possible
towards the presheaves on any Topos.6 Therefore, by this construction, every Topos (typically a pre-sheaves category, but I
shall come back to this) can play an analogous relativizing
role, without for all that becoming an indispensable absolute.7
that Zalamea does more explicitly. Among the interesting analyses of the
contemporary mathematical work that these volumes present, I want to
single out the articles by McLarty on the notion of scheme (a topological
space with a sheaf of rings or more), and of Urquhart on mathematical
inventiveness in physics, often non-rigorous or presenting an original
informal rigour, a co-constitution of sense and therefore, gradually, of
new mathematical structures (see Mancosu 2008).
6 One of the few required properties is the locally small hypothesis:
every collection of morphisms Hom(A,B), must be a set (see Mac Lane
and Moerdijk 1992). Once more, a close look at the assumptions and
the proof (its right level of generality, in this case) is essential for the
understanding of a theorem.
7 Many (all?) categorial objects can be codified as sets, even Set, the para-

215

Speculations VI
Similarly, Maddy identifies mathematical practice with the
work done upon a structureless set theory and identifies, in
this non-structured assembling of points and elements, the
cognitive foundations of mathematics. These approaches
are in explicit contrast with the key ideas of Zalameas book
which, centered upon categorical universes of geometrical
inspiration, attempts to make us appreciate the structural
sense of mathematical construction.
Luckily, soon afterwards, a reference to Chtelet enlightens
us with a much different insight. References (perhaps too
cursory) to that masterpiece that is Chtelet 1993, bring our
attention back to the gesture constitutive of mathematical
objectivity, which lies on the border of the virtual and the
actual, in a tight interrelation between the construction of
objects of study and objectivity in physics and the analysis
of the organizational structures of the world, starting with
symmetries. Chtelets book, it should be emphasized, is also
an history; rather, it is a historico-rational reconstruction
of the rich entanglement between physics and mathematics running through the 1800s up to, and stopping short of,
the advent of Set Theory. Regarding some related aspects of
contemporary mathematics, Patras 2001 (a book that Zalamea cursorily mentions), has retrieved the point of view of
structural mathematics with a philosophical competence
rare to find in a mathematician. Patras exhibits the weaving together of structures and transformations that governs
mathematical construction from the inside, from the point
of view of mathematical practice and invention.
In general, the origin of meaning in mathematics is to be
found in the ways in which it allows us to organize, to strucdoxical set of all sets. In every such occasion an ad hoc construction or
codification is necessary, and in such a case, we pay the price of stretching the sets, up to cardinals as inaccessible (Kanamori 2003) as they are
far from the construction one wants to interpret. These are codifications
that push the meaning of categorial structures out of sight. The point,
indeed, is not the possibility of a coding, perhaps a meaningless one: it
is rather the relativizing -- and geometrical diagrammatical knowing
proper of categories, which is sensitive to coding, as we might put it,
that makes all the difference.

216

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
ture, the world. Only then does it detach itself from the world
in the autonomy of constitutive gestures, between the virtual
and the actual where, at a farther remove from the original
constitution of meaning, one obtains relevant results at the
intersection between constructions of diverse origin. From
classic algebraic geometry and differential geometry, two very
productive blends, to sheaf-cohomology and cohomologysheaves, between complex analysis and algebra (179), where,
as Serre puts it, such problems are not group theory, nor
topology, nor number theory: they are just mathematics.
Structural continuity becomes conceptual continuity, a
navigation between concepts as a sophisticated technical
transits over a continuous conceptual ground.
In brief, the study of structures, of their continuous
enchanements and deformations, is an essential component of foundational analysis; without it one can at best
hope to do Set Theory.8 The latter is an extremely interesting
theory and category: the error is to make an absolute out of
it and to posit sets of meaningless points at the root of every
mathematical construction, in what amounts to a ruinous
disintegration of sense. The origin of mathematics and its
principle of construction are located in that which is meaningful, in thought operations that structure and organize the
world, but which then go to intersect on planes far removed
from the world and acquire by these conceptual interactions
a proper mathematical sense.
Thus Zalamea cites the Langlands Program. Langlands
dared to write to the more famous Andr Weil proposing an
extensive web of conjectures by which number theory, algebra,
and analysis are interrelated in a precise manner, eliminating
the official divisions between the subdisciplines, and suggesting that one approach the world of the complex variable
and the world of algebraic extensions functorially, by way of
8 Consider that the axioms of Set Theory, essentially created in order to
adjudicate the validity of principles of well-ordering and choice, are
silent on them: a failure for a whole program. A refined analysis has been
conducted, in structured environments wherein these constructions can
be relativized, by Blass (1983).

217

Speculations VI
group actions. This will indicate an unexpected equivalence
between certain differentiable structures associated with an
extended modularity (the automorphic forms associated with
the linear group) and certain arithmetical structures associated with analytic continuations (the L-representations of the
Galois group) (180-182). Here we see groups again, and thus
transformations and symmetries, both technical and conceptual, which allow for this splendid structural unity which lies
at the heart of mathematics: in a certain sense, Langlands
program extends Erlangens program to Number Theory. So
technical and conceptual invariants get transformed, like the
generalized analysis of continuity that underlies the notion
of fibration, and the subtle interplay between continuous
and discrete, the founding aporia of mathematics [] that
drives the discipline, as Thom puts it (138).
Zalamea recognizes that nothing could therefore be further
from an understanding of mathematical invention than a
philosophical posture that tries to mimic the set-theoretical
analytic, and presumes to indulge in such antiseptic procedures as the elimination of the inevitable contradictions
of doing mathematics or the reduction of the continuous/
discrete dialectic (183-184). This, I would add, extends all the
way to the discrete-computational approaches, flat (or better:
unidimensional) visions of the world, according to which
the Universe (Wolfram and others), the brain (too many to
mention), or DNA (Monod, Jacob, Crick) would be a (large,
medium or small ) Turing Machine (see Longo 2009, 2012).
The great invention of Gdel, Turing and others in the 1930s,
the theory of logical-formal - computability, instantiated in
machines that today are changing our world, is projected by
these stances to the world and identified with it, even while it
was originally developed, within (Frege and) Hilberts logical
systems, thus to explicitly distinguish itself from the world.
Nowadays these approaches are not so counterproductive in
physics, where they are mostly ignored: in biology, instead,
such frameworks and methods exclusively grounded on
discrete sets of strings of code have profoundly impaired the
comprehension of biological phenomena. It is here that I will
218

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
introduce a correlation of outlooks, the necessity of which I
hope to convince the reader of.
Let us begin with an example. The discrete-computational
outlook has not helped us (or has not permitted us) to detect
the role of endocrine perturbators of the 80.000 (sic) artificial
molecules that we produced in the twentieth century. These
were mostly presumed to be innocuous, below arbitrarily
imposed individual thresholds, since not stereo-specific
(not in exact physico-chemical-geometric correspondence)
and thus unable to interfere with molecular-computational
cascades, necessarily stereo-specific, going from DNA to
RNA to proteins (the Central Dogma of molecular biology),
and with hormonal pathways. It should be noted, indeed,
that exact molecular stereo-specificity was deduced, against
experimental evidence that were already available (since 1957,
see Elowitz and Levine 2002; Raj and Oudernaaden 2008): it
is necessary, as Monod (1972) puts it, for the transmission of
computational information and for the genetic programme
to function. Thus, negating the role of context in genetic
expression and hormonal control, the consequences (direct
and indirect) of the finite combinations of said 80.000 molecules on the organism and on the chemical ecosystem of the
living have receded from view. Cancer incidence has grown
in the last half century, across all age groups, jointly to the
halving (sic) of the average density of human spermatozoa
in Western countries (Diamanti-Kandarakis et al. 2009; Soto
and Sonnenschein 1999, 2010). As for cancer, the failure of the
fifty years old, DNA centered, molecular approach has been
recently aknowledged even by one if its founding fathers,
Weinberg (2014).
In contrast with the claims of the informational analyses,
macromolecular interactions even within the cell, where the
macromolecules in Brownian motion have quasi-chaotic entalpic oscillation are stochastic, and are given as probabilities,
and these probabilities depend upon the context; a strongly
influential context, made of interactions, deformations, morphogenetic fields, biological networks and structures, and so
on (see Elowits and Levine 2002; Noble 2006, among others.
219

Speculations VI
See also Longo and Montvil 2014). A context, then, made of
ecosystemic structures and their transformations, very different from the fragmentation of the analysis of organisms as
sets of molecules promoted by the still-dominant Laplacean
reconstruction (a linear one, molecule after molecule, a cartesian mechanisms says Monod).
The discourse on the foundations of mathematics has
played an enormous scientific, suggestive and metaphorical
role in these events: the absolute certainty of the arithmetical discrete/finite, decidable (and thus programmable) has
produced, on the one hand, original and powerful machines,
perfectly artificial instruments for formal calculus allowing
the networking of the world, while on the other it has contaminated our worldview even though, originally, it had been
lucidly and courageously originally proposed, by Frege and
Hilbert, in order to detach those foundations from the world.
Logics, Topos, and Symmetries. In Brief.
Returning to less dramatic topics, another author Zalamea
often refers to is Lawvere. The latter transferred Grothendiecks notions into an original analysis of Logic, grasping
how Topos Theory and, more generally, Category Theory
presents a permanent back-and-forth between the three
basic dimensions of the semiotic, emphasizing translations and pragmatic correlations (functorial comparisons,
adjunctions) over both semantic aspects (canonical classes
of models) and syntactic ones (orderings of types) (191).
Going back to my first scientific life, I remember the interest
around the categorical interpretation of Type Theory, which
owes much to many brilliant mathematicians who Zalamea
has no space to mention (but who are cited in Longo 1988;
Asperti and Longo 1991). A wonderful community, where a
logical sensibility and I am thinking of the challenge offered by Girards Impredicative Theories of Types found
in categorical semantics a strong link to the mathematics of
structures that concerns Zalamea. The crucial point is the
geometrization of logic and its relativization to Topoi
220

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
that can have different internal logics, properly correlated
by functors and natural transformations.
In these circles, Fregean quantifiers, for example, are interpreted in terms of adjunctions. More precisely, existential
and universal quantifiers become right and left adjuncts to a
sort of diagonal functor: the pullback along a projection. Then
the existential quantifier is interpreted as the projection in
a product of objects in well-defined Topoi, and the universal
quantifier is its dual, modulo an adjunction. So the level of
effectivity of the existential quantifier (the possibility of
effectively constructing the mathematical object whose
existence is predicated), a delicate issue that has been the
object of a century-long debate, is relativized to the effective
nature of morphisms in the intended Topos as a (relative)
Universe that is, to its internal logic. The meaning of
logico-formal construction, then, is given by a reflexive interplay of invariances and symmetries (the duality present
in an adjunction) without the need for an understanding of
for every as meaning for every, or that exists really means
exists just as, for far too long, we have been told that snow
is white is true just when snow is white, a truly remarkable
mathematical discovery. When the geometric meaning of
an adjunction is known, qua profound and omnipervasive
construct of Category Theory, the meaning and the relation
between the quantifiers is enriched with a new structural significance through the construction described above. That is,
they become immersed in a geometric context, a universe of
dynamic and modifiable structures. In particular, it becomes
possible to go from one logic to another, from one Topos to
another, studying their invariants and transformations, that
is, the functorial immersions and the adjunctions correlating them. For this reason I often say, in provocative manner,
that I am happy to leave the question of truth to priests and
analytic philosophers: we operate constructions of sense, we
organize the world by proposing and correlating structures
that have a meaning because of our being world-bound active humans in different conceptual worlds which we strive
to put into dialogue. Let us not confuse this with the fact that
221

Speculations VI
the judge seeks, in witnesses for example, the truth: science
is not a testimony of, but an action upon the world, aimed at
organizing it and giving meaning to it.
I will return shortly to this extremely timely geometrization
of Logic, a royal way out of the narrow singlemindedness
of the logico-linguistic turn. In this regard, Zalamea quotes
Girard who, within Proof Theory, has subsumed the same
structural sensibility, the same distance from Tarskian truth
and its ontological flavours. I remember when I first attended,
in the 1980s, a talk by Girard on Linear Logic; I asked him
why, after having radically modified the structural rules
of logics, changing their symmetries in formal notation, he
had introduced a certain inference rule. He replied: for reasons of symmetry.9. Symmetries are at the core of the close
relationship between physics and mathematics, ever since
Archimedes asked himself: why doesnt a scale with equal
weights on both sides move? And answered: For reasons of
symmetry. Guided by the same symmetry reasons, Sacharov
and Feynman proposed anti-matter, thus giving a meaning
faced with experimental phenomena in need of explanation to the negative solution of Diracs electron equation.
Alas, unfortunately (or fortunately?) cellular reproduction
is at the heart of ontogenesis and phylogenesis, also because
it is asymmetrical.
More on Invariance and Symmetries,
in Mathematics and the Natural Sciences
1-Between mathematics and physics: Symmetries,
Gestures, and Measures.
I have been too critical, much more than Zalamea is, of
Set Theory as a foundational discipline, since there is one
9 Symmetry principles or more precisely principles of inversion were
already present in Grentzens sequent calculus, to which Girard explicitly
refers to. They permit the generation of a calculus starting with logical connectives, and to finely analyze the properties of proof-theoretic
normalization (see Negri and von Plato 2001).

222

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
concept about which it has been the field of a rigorous and
useful foundational analysis: the question of the infinite.10
This is a crucial concept in mathematics. All mathematics
is construction to the limit, starting with the line with no
thickness of Greek geometry, a limit construction, all the
way to the higher constructs I have discussed above. It has
come into relation with physics since Galileos asymptotic
principle of inertia. Great merit goes to Shelah, whose work
Zalamea discusses at great length, for he demonstrated that
the theory of singular cardinals corresponds to the idea of
seeking natural algebraic invariants (homotopies, homologies)
for topology (202). From there, we are referred to Serres work
on homotopy, which makes possible an algebraic-topological
relativization of the notions of finite and infinite. Once again,
it is a relativizing operation, breaking with the absolutes of
logicist formalisms, according to which the finite is locus of
certainty and absoluteness. Likewise, in physics, the Riemann
Sphere, a bidimensional model of the relativistic universe, is
infinite for its surface-bound inhabitant moving towards the
poles, whose meter stick progressively contracts; it is finite
as observed from an external reference frame.
At the level of groups, however, a discrete combinatorics can
be fundamental; indeed, Zalamea refers to the GrothendieckTeichmller groups, which may come to govern certain
correlations between the universal constants of physics (the
speed of light, the Planck constant, the gravitational constant),
while, conversely, certain mathematical theories originating
in quantum mechanics (non-commutative geometry) may
help to resolve difficult problems in arithmetic (the Riemann
hypothesis) (205). As Zalamea tells us, here we witness absolutely unanticipated results, which bring together the most
10This analysis extends all the way to the recent and daring anti-Cantorian
explorations of Benci, Di Nasso and Forti (in Blass et al. 2012). According to them, as for Euclid, the whole is larger than its parts, even for
infinite sets (at least when denumerable: this approach, for the time
being, is not extended beyond the denumerable. For this latter domain,
we will probably have to look beyond the category of sets, towards other
structural invariants).

223

Speculations VI
abstract mathematical inventions and the most concrete
physical universe (206).
Through a back-and-forth between mathematics and physics,
various intersections far from the world are drawn out, between domains with roots in diverse conceptual constructions,
each originating in different organizational actions upon the
physical world. It is neither unreasonable nor surprising that
the locus of conceptual invariance and of the analysis of its
transformations mathematics should influence theoretical
physics. Beyond the strict relation mentioned above between
mathematical symmetries and conservation principles in
physics (Noether, Weyl), the physicists theoretical work begins
from the invention of appropriate, and very abstract, mathematical phase-spaces (observables and pertinent parameters)
like the spaces of state-function in quantum mechanics or
Hilbert spaces; all phase-spaces the physicist uses or builds
to analyze (generic) objects and (specific) trajectories, result,
in turn, from symmetries and invariances. I will try to sum
up here analyses and notions which are central to attempts to
differentiate and establish a dialogue between mathematics,
physics, and biology (as exposed in Bailly and Longo 2011
and Longo and Montvil 2014).
Mathematics and physics share a common construction
insofar as they isolate and draw pertinent objects, perfectly
abstract and with pure contours like Euclids lines with no
thickness, edges of figures drawn on the veil of phenomenality, at the interface between us and the world. Euclid, indeed,
invents the difficult notion of border: his figures are nothing
but borders, and thus without thickness one thinks of Thoms
cobordism (Rudyak 2008). These objects, in mathematics as
in physics, are generic, that is interchangeable, symmetrical
according to permutations within their definitional domains.
A right-angled triangle in Euclid, a Banach space, or a sheaf,
are all generic, as are Galileos weight, an electron, a photon,
and so on. These are generic insofar as they are invariants of
theory and of physical experience, symmetrically permutable
with any other. So that the same theory can deal with falling
apples and planets as generic gravitational objects, just as the
224

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
even more marked theoretical invariance of the theory of
relativistic bodies allows us to unify gravitation and inertia.
The genericity of objects and of structures, therefore, is the
result of a fundamental symmetry/invariance, shared by both
mathematics and physics.
Beginning with the genericity of its objects, physics analyzes
trajectories in a suitable phase-space. The classical one based
on momentum and position (or energy and time) is only one
among many (thermodynamics, for example, operates within
a space defined by pressure volume temperature, and has
added a revolutionary observable: entropy). These trajectories
are specific, unique, and are imposed by the geodetic principle
in its various instances. Even in quantum mechanics, where
the quanta certainly do not follow trajectories in space-time,
the Hamiltonian allows the derivation of the Schrdinger
equation, defining the trajectory of a probability amplitude
in Hilbert space. But the Hamiltonian, or the extremization
of a Lagrangian functional, follow from a conservation principle a principle of symmetry as Noethers theorems have
explained (see Kosmann-Schwarzbach 2004; Bailly and Longo
2011). Here is the extraordinary unity, completely construed
or better co-construed, of the physical-mathematical edifice.
Here is the power of its intelligibility, utterly human, for we
animals characterized by a fundamental bilateral symmetry
who, in language and intersubjective practices, organize the
world, our arts, and our knowledge in terms of symmetries
(see Weyl 1949, 1952, followed by Van Fraassen 1993) and,
subsequently, their breaking.
Such unity will be discovered in the symmetry breaking
constituted by the non-Euclidean modifications of Euclids
fifth postulate which yields the closure of the Euclidean
plane under the group of homotheties a breaking that will
allow Einstein to give a mathematical foundation to relativist
physics, beginning with the astonishing measurement of the
invariance of the speed of light. Likewise, in Connes noncommutative geometry, which includes physical measure in
the foundations of his approach: Heisenbergs matrix algebras,
from which it derives in analogy with Geflands construction,
225

Speculations VI
are built starting with the non-commutable nature of quantum measurement. In a striking difference from arithmetical
foundations, geometry, the privileged locus of invariance and
transformations, has always had an origin in a constructive
relationship of access to space and its processes: from the
Greek compass and straightedge to Riemanns rigid body, to
the algebras derived from Conness quantum measurement,
yet another bridge between mathematics and the universe
of physics.
To sum up, a fundamental component of the unity we have
delineated between mathematics and the theorization of the
inert is this central role assigned to the genericity of objects
and the specificity of their trajectories, both being definable
in terms of symmetries. To this we should add an active relation to the world, grounded on both the constitutive gesture
of the continuous line, of the trajectory a movement at the
origin of the phenomenic continuum and on the access to
the world as mediated by measurement: classic, relativistic,
and quantum. Following Zalamea, I will return, in what follows,
to some contemporary consequences of these considerations
(which sum up ideas extensively developped in Bailly and
Longo 2011 and in Longo and Montvil 2014, and are directed
towards a discussion of biology).
2- What About Biology?
What can we say about the theorization of the living? The
only great biological theory, Darwins, was born by positing
some principles: of which the first in particular, descent
with modification (indispensable for the second, selection), stands in stark contrast to those conservation principles (symmetries) which, starting with Galileos inertia
and the geodetic principle (think of Hamiltons variational
method, contemporary to Darwin), were taking center stage
in physics. Descent with modification is a principle of
non-conservation of the phenotype, of organisms, of species
and of all the observables of Evolutionary Theory. The morphogenetic iterationin the living, in particular reproduction
226

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
as conservation by inheritance, is never identical to itself,
and this must be take its place as a fundamental principle,
together with Darwins, of the intelligibility of ontogenesis
(see Longo et al. 2014).
We are working towards an understanding of onto-phylogenetic trajectories as cascades of symmetry change, a
kind of extended critical transitions (see below), borrowing a method from physics: a mathematical construction
of objectivity, yet with dual principles. Critical transitions
capture the continuity of change that is proper to reproduction. The challenge is to unify ontogenesis and phylogenesis,
on the basis of the same, or similar, principles (see Longo et
al. 2014), thus towards a theory of organism and therefore
of ontogenesis, avoiding the prescientific metaphors of an
Aristotelian homunculus codified in the DNA (even when
the defenders of such theories dress their ideas in modern
garments: the homunculus is in a machine code and the
DNA contains both the program and the operating system
[Danchin 2009]).
The problem is that biological trajectories, cascades of
changes of symmetry in constant interaction with the ecosystem, must be considered as generic: they are possible
trajectories among the many which are compatible with the
ecosystem the limbs of an elephant, of a kangaroo, of a whale
(its vestigial forms) are so many possible evolutions originating form a same tetrapod vertebrate. Whats particularly
hard to grasp is that they are possibilia in phase-spaces (to
use a physics jargon), not pre-given but rather co-constituted
with trajectories: so an organism, in phylogenesis as well as
in ontogenesis, co-constructs its ecosystem: consider how,
two to three billion years ago, bacteria created oxygen, beginning with a primitive atmosphere which contained none or
in negligible amounts. And so the pertinent observables
that is, the phenotypes are modified up to speciation. The
result of this evolutionary trajectory is an historical and
individuated object, a specific organism, the result of a contingent cascade of change of symmetry (qua changes of the
coherence between organism and ecosystem) channeled by
227

Speculations VI
massive historical constraints. One of the most important
of which is the DNA: the imposing chemical trace of an history, continuously employed by the organism throughout
the course of ontogenesis.
To sum up: biological trajectories are generic, while their
objects are specific a radical duality, as opposed to the
physical-mathematical realm, where we pointed out the genericity of the objects and the specificity of the trajectories.
Such duality profoundly modifies the role so rich in physics of symmetries, invariances and transformation. To the
impenitent reductionist, hellbent on an abstract physics (and
not the physics of the historically-situated theories) to which
everything must be reduced, we respond (see the introduction of Longo and Montvil 2014) with a recommendation,
for example, to try to reduce the classical domain to the
quantum one, or the hydrodynamics of incompressible fluids
in a continuum to quantum mechanical principles, if she can
after all, there are both classical and quantum dynamics (and
plenty of water) at play within a cell. The unity of knowledge
and of its scientific instruments, starting with unity in physics, is a hard-won conquest as in the case of quantum and
relativistic physics and not a theoretical a priori.
I mention these problems both because they are my current interests and because the construction of objects and
structures in mathematics has proceeded in lockstep with a
prodigious construction of objectivity in physics, simoultaneously locating in the richness of language and of historically
located human gestures an autonomy that pushed it steadily
away from physical experience (where is Euclids thickless
line to be found? Where is a Grothendieck pre-sheaf located?).
And yet, considering the analogous approach in physics and
mathematics to objects and trajectories, this was a process
of constitution capable of falling back again upon physics,
through unexpected avenues: think of the marvelous story
of Cardanos imaginary numbers, having an highly abstract
algebraic origin and yet being today essential to talk about
microphysics (yet Argands and Gausss interpretation allows
us to discern a possible role for them in the description of
228

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
wave amplitudes and their trajectories: before falling back
upon the world, they became a rich geometric structure).
This parallel construction of objects and concepts does not
merely concern the interaction of physics and mathematics.
Indeed, even in the ambit of proof, mathematics does not
proceed by way of demonstrations of already-given formulae as the formalist caricature would have it and physics
does not construct theories as summations of experiences
and facts. Neither proofs nor theories are already there, not
even in the most dynamical and weakly-Platonic sense. The
construction of sense plays a powerful role in proof, even
arithmetical proof (see Longo 2010, 2011); likewise, physical theory tells us which observables are to be isolated and
analyzed, which experiences to have, which phenomena to
observe. Mathematics and physics are the result of a laborious effort of knowledge construction, as Weyl has it, through
a non-arbitrary friction with the world. Non-arbitrary and
effective precisely because rich in history and contingency:
mathematics and physics are thus a human praxis in and
towards the world, as Peirce a thinker Zalamea often likes
to refer to would say.
Contemporary biology poses enormous challenges: to face
them we would need to combine the imagination of Newton
(a Newton of the blade of grass, as Kant has it, without denying
the possibility of such a science), with his differential calculus
as infinitary construction to understand the movement of
the finite; of Hamilton, with the variational method for the
geodetic principle; of Dirac, with his delta, for a long time
without any mathematical sense; and of Feynman, with his
integral, the solution of a still-non-defined equation. The
principal invariant in biology (fortunately not the only one)
is variability: it allows diversity adaptability, at the heart of
the structural stability of the living. What to do with our
invariantentheorien?

229

Speculations VI
Groups Everywhere, Metrics Everywhere
Among the omnipresent references to Grothendieck, Zalamea underlines time and again how his work incorporates
a transit between objects (variations, perturbations) so as
to then proceed to determine certain partial stabilities (invariants) beneath the transit (212). As for the invariants, I
have often referred, as Zalamea does, to those correlated with
symmetries, i.e. group structures. But together with groups (to
be interpreted as instruments of action upon spaces, all the
way to the most abstract ones due to Grothendieck), a crucial
epistemic role should be assigned to semigroup structures.
As it is observed in Bailly and Longo 2011, on the one hand
we should consider the gnoseological and mathematical
complex of {space, group, equivalence relation}, on the other
that of {time, semigroup, ordering relation}. In the passage
between the two we see a useful instrument to analyze the
interplay between space and time in the natural sciences, as
well as the difference between physics and biology: oriented/
ordered time plays a crucial operatorial role in biology, as we
say also in Longo and Montvil 2014, well beyond its role as
parameter in physics. In this regard, Zalamea insists on the
role of semigroups in the hyperbolic varieties of Lax and
Phillips (218). These are collections of operators Z(t), with a
parameter that can be interpreted as time, which permit the
construction of the deep connection that lets us unfold the
intrinsic meaning hidden in differential equations like the
non-euclidean wave equation, a meaning that can be glimpsed
precisely in virtue of the semigroup Z(t) (220).
In this inexhaustible search for unity, not forced towards
impossible reductions, but constructed with bridges, correlations, and structural passages, we can naturally mediate
between the Poincar plane, seen as a non-Euclidean model,
with its differential Riemannian geometry and analytic
invariants, on the one hand; and the same plane, seen as a
complex model, with its theory of automorphic functions and
arithmetical invariants, on the other (220). Here we arrive
at Conness programme for non-commutative geometry, a
230

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
programme for the reconstruction of vast sections of mathematics, grounded on the non-commutativity of quantum
measure (and its algebras). The objective of this geometrization
of quantum mechanics is to contribute to its intelligibility
and, ultimately, to deliver a unification with the relativistic
universe, radically changing the theory of space not a mere
background, as string theorists claim. Zalamea adroitly sums
up several bridging aspects, correlating them with the work
of other geometers, starting with the recent developments
of Riemaniann differential geometry, with particular focus
on the passage from infinitesimal manifolds (Riemann) to
C*-algebras of compact operators (Hilbert, von Neumann),
the passage from dual K-homology (Atiyah, Brown, Douglas,
Filmore) to non-commutative C*-algebras (Connes), the passage from the index theorem (Atiyah, Singer) to the handling
of non-commutative convolutions in groupoids (Connes),
the passage from the groups and algebras of modern differential geometry (Lie) to quantum groups and Hopf algebras,
the passage from set-theoretic punctuality to the actions of
non-commutative monoids in Grothendieck topoi, etc. (224).
There is no doubt in my mind that this allows for a correspondence in fieri between mathematics (as a study of
quantities and organized in structures) and the cosmos (as
order), as Zalamea argues, legitimately philosophizing from
a conjecture of Cartier. But this shouldnt be considered a
new Pythagoreanism, in my view: it is we who single out
elements of order in the cosmos (those we can and want to
see symmetries for example). As Kontsevich, quoted by
Zalamea, has it, in physics we begin with very little: where
one doesnt see structures so much as the symmetry, locality
and linearity of observable quantities (229). We then enlarge
these almost Gestaltic elements (symmetries and locality), we
generalize them, and we transform them into the language of
a metaphysics-rich communicating community. Finally we
project them back again upon the cosmos, recognizing it as
orderly because intelligible, and intelligible because orderly.
This process is legitimate because, in this theoretical backand-forth, our friction and action upon the world are real:
231

Speculations VI
the world resists, it says no, and channels our epistemic
praxis, which is of an eminently organizational character,
and it is always active.
Such knowledge construction works because of this cognitive entanglement, beginning with the common genericity
(of objects) and specificity (of trajectories), both physical
and mathematical: the first brick of an enormous physicomathematical edifice of our making. No surprise then, a
surprise still affecting Kontsevich and Zalamea; we are left
with great admiration for such a majestic, but very reasonable, mathematical construction. Similarly, the linguist is
not surprised if, when we talk, we understand each other:
language was born with dialogue, through the practice of
mutual understanding and communication. The linguist
surely admires a great poem which, with words, introduces a
different worldview or an original intelligibility of humans,
without ontological miracles but merely with the strength of
the words meaning, a co-constituted product of our human
community. Alongside myths, poems and tragedies rich
in human experience, in human, concrete and lived praxis
as well as in metaphysics we have been able to propose the
structures of mathematics with their invariants and transformations, rich in those glances and gestures which organize
the real, as well as rich in metaphysical nuance starting with
Euclids line, a limit notion resulting from a dialogue with
the Gods. Mathematics is written in natural language, it is
a language and a gaze upon the world, at and from the limit
of the world (mathematics is the science of the infinite as
Weyl [1932] writes).
However, we only see perspectives, albeit coherent and
profound ones; points of view on fragments of the world,
we organize and make accessible small corners of it. And as
soon as that small (but oh so important) brick concerning
physico-mathematical genericity and specificity is removed,
as happens in the analysis of the living, we find ourselves in
trouble. Yet it is nothing unsurmountable: we just have to
work on it with the same freedom and secular independence
of thought, action, construction and exchange proper of the
232

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
founding fathers of the physico-mathematical, abandoning
the ambition of finding the theoretical or mathematical
answer already there, written by God in the language of
already-existing mathematized physics.
Referring to Peirce, Zalamea too highlights the progressive
constituting of knowledge of the world:
we see how the world consists in a series of data/structures (Peircean
firstness), registers/models (Peircean secondness) and transits/functors (Peircean thirdness), whose progressive interlacing into a web not
only allows us to better understand the world, but which constitutes it
in its very emergence. (237)

The important thing is to break out, even in foundational


analysis, from an absolute mathematics, a mathematics
at rest, in the style of Russell and proceed towards a relative mathematics, a mathematics in motion, in the style of
Grothendieck (240). The entire work of contemporary
mathematics, carefully recounted by Zalamea, aimed at the
production of
remarkable invariants without any need of being anchored in an absolute
ground. We will therefore take up a revolutionary conception which has
surfaced in contemporary mathematics in a theorematic manner: the
register of universals capable of unmooring themselves from any primordial absolute, relative universals regulating the flow of knowledge. (242)

Developing the theme of relative universals, Zalamea


introduces Freyds allegories: abstract categories of relations, exposed in diagrammatic terms via representations
that obviously a functional, set-theoretic reading would fail
to detect (243). I want to stress that, in general, categorial
diagrams are not equivalent to the equations to which they
can be formally reduced: the diagrams indeed highlight symmetries that are merely implicit, invisible, in the equations;
they need extracting, just as Noethers theorems extract
symmetries from the equations of physics.

233

Speculations VI
Freyd shows how, starting from pure type theories with certain structural
properties (regularity, coherence, first-order, higher-order), one can
uniformly construct, by means of a controlled architectonic hierarchy,
free categories that reflect the given structural properties in an origin
(regular categories, pre-logoi, logoi, and topoi). (243)

In this way, all the invariants of logico-relational transformations beyond the particular variants of any specific
logico-mathematical domain are expressed in a maximally
synthetic and abstract way. As usual, the analysis of transformations, of preserved structural invariants, and of variants
(which can however have a local sense) is at the heart of
mathematics, and this is confirmed by the logical-foundational
spirit of Freyds work. Referring to the latter, and taking his
moves from the Yoneda Lemma, Zalamea uses the occasion
to explain, as I mentioned above, that pre-sheaves categories
can be considered as the general locus of the continuity
wherein every discrete category can be embedded. Like Thom,
one comes to the conclusion that the continuum underlies
(is an archetype) for the discrete as well (Thom argues that
a discrete set is nothing but a collection of singularities in
a continuum).
Without necessarily according ontological priority to the
one or the other, I would like to observe that, in the natural
sciences, the discrete and the continuum organize the world
differently, and this can be demonstrated: by analyzing the
different role of symmetries and their breakings, which these
mathematical structures, when employed for theoretical organization or simulation, accentuate and project upon physical
and biological processes (see Longo and Montvil 2014a).
Having passed through a technically pertinent close-up of
the reverse mathematics of Friedman and Simpson, Zalamea
demonstrates how the work of Zilber contributed to giving a
Grothendieckian understanding of the model theory of Tarskian tradition (Chang, Keisler): no more logic + universal
algebra but algebraic geometry + fields (Shelah, Hrushovski,
Zilber, Hodges). With Zilber we have the emergence of groups
everywhere invisible at first, but lying in the depths (ar234

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
chetypes) (256). A kind of renaissance and generalization
of Erlangens program, as Zalamea rightly notes.
An analogous motto allows us to grasp a central element
of Gromovs contribution to geometry: smoothing and
globalization that are tied to the notion of metrics everywhere (259). Then Zalamea hints, with fine synthetic and
analytic skill (that is, with great command of language and
pertinent mathematical references, as always), to the work
of Gromov on partial differential relations, on symplectic
varieties, and on hyperbolic groups (259) a work enriched
by a certain sensitivity, proper of the French-Russian school,
to the play between geometric insight, analytic virtuosism
and physical applicability. Introducing pseudoholomorphic
curves and seeking the
invariants of those curves, Gromov shows that the spaces modulo the
curves are compact, and that it is therefore possible to work out a natural
theory of homology, which leads to the Gromov-Witten invariants; in the
last instance, the new invariants allow us, on the one hand, to distinguish
an entire series of hitherto unclassifiable symplectic varieties, and,
on the other, help to model unexpected aspects of string theory. (262)

Once again, the analysis of the invariants and the transformations preserving them relativizing the movement
between a structure to another is at the core of Gromovs
work on Riemaniann manifolds, within a program of geometrical group theory described as the project aiming at
characteriz[ing] finitely generated groups, modulo quasiisometries, which is to say, modulo infinitesimal deformations of Lipschitz-type distances (264).
In Chapter 8, Zalamea synthetizes some of the themes
touched in the book, in order to propose his own vision of
a transitory ontology. It is a relativizing, yet not relativist
vision (of either the weak or the anything goes variety),
an Einstenian vs. Newtonian one, at the center of which lie
transformations (passages, transits) and pertinent invariants:
the transit of mathematical objects consists in finding suitable invariants (no longer elementary or classical) behind that
235

Speculations VI
transit (271). And so Zalamea himself sums up the themes
he examined more extensively earlier in the book
motifs [p.144-146], pcf theory [p.201-202], intermediate allegories
[p.245-246], Zilbers extended alternative [p.257], the h-principle
[p.263], etc. [] neither absolute foundations nor fixed objects, not
everything turns out to be comparable or equivalent, and where we can
calculate correlative archeal structures that is, invariants with respect
to a given context and a given series of correlations which, precisely,
allow differences to be detected and reintegrated. (272)

Representation theorems,which Zalamea often mentions


in his book, assign a key role to strong and diverse specifications of the notion of group. To emphasise this role, I borrow
Zalameas own list of topics (specifying, in square brackets,
where each theme has been considered), always examined
with a refined informality that manages to be both complete
and informative.
homology and cohomology groups [p. 142-148, 178-179], Galois groups [p.
150, 155, 225], group actions [p. 162-163, 180-181], Abelian groups [p. 165],
homotopy groups [p. 176], algebraic groups [p. 184], the GrothendieckTeichmuller group [p. 225, 233], Lie groups [p. 223], quantum groups [p.
223], Zilber groups [p. 255-256], hyperbolic groups [p. 264], etc. (272)

This demonstrates a dynamics of webs incessantly evolving as they connect with new universes of mathematical
interpretation. [] This just goes to reinforce the position
of Cavaills, who understood mathematics as gesture (273).
Such are organizational gestures of correlated mathematical
universes, correlated by a web of transformations, like the hand
gesture that organizes space, gathers, delimits, and transfers,
as we can say with Chtelet. This process assumes an historicity that serves to highlight the sense and the relationship of
mathematics vis--vis the real: mathematics works (where it
does work) and has meaning because it is constituted through
a human all too human praxis. All too human because it
is anchored to pre-human invariants, those of our actions in
236

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
space and time; universal, for us historical and speaking human beings, precisely because pre-linguistic and pre-historical,
even though language alone allowed the transformation of
practical invariants into concepts and structures. And, in
language, writing, as Husserl (1970) observed, has further
contributed to the process of the stabilization of concepts.
Considering the correlations between groups, symmetries
and invariants, in the context of this section on groups everywhere, metrics everywhere, I would like to mention the
role of (animal) memory in the constitution of invariants.
Memory is forgetful, that is one of its essential properties:
we, as animals, forget irrelevant details of an action, of a lived
experience. Irrelevant, that is, with respect to the protensive
intentional (conscious) or not gesture, already done or
still to be performed: memory is selective in both its constitution and in its re-activation. This selective choice allows us
to undertake once again a given action in a similar but not
identical context, to operate another protension or prevision,
counting on the relative stability of the world, through changing distances, for example, which we attempt to organize in
stable metric evaluations. We do not access memory as we
would access a digital hard drive. The protensive gesture, I say
with and beyond Cavaills, reactivates memory every time:
not in a passive way, but choosing, selecting and constituting
new practical invariances, beyond those isolated and selected
by memory in its constituting process. Animal memory is
reactivated in a protensive manner, or better, it is re-lived for
a purpose, be it a conscious or non-conscious one, forgetting
all that is irrelevant to the present goal:(Edelman and Tononi
2000) argue that, in the act of memory, he brain puts itself
in a lived state..
Meaning derives, moreover, from the intentionality, even
a pre-conscious one, that inheres in protensive gestures,
particularly in a perturbative modality. It is that which
interferes with, and which operates a friction upon, the protensive action which acquires, for us as animals, a meaning.
And there is no protension without retention. Obviously,
then, a digital machine with a perfect memory cannot do
237

Speculations VI
mathematics, because it cannot constitute invariants and
its associated transformation groups, because a perfect,
non-protensive memory does not construct meaning, not
even mathematical meaning. At most, the machine can help
with formal fragments of proofs, or check, a posteriori, the
formalized proof, or parts of it (proof-assistance and proofchecking are burgeoning fields). Only animal memory and
its human meaning allow not only the construction of concepts and structures, but proof as well, as soon as the latter
requires us to propose new concepts and structures, or the
employment of ordering or invariance properties which go
beyond the given formal system (well-ordering, say, or the
genericity of infinitary structures). It is thus that recent results on the concrete incompleteness of formal systems can
be interpreted: meaning demonstrably lurks in the proofs of
formally unprovable theorems (see Longo 2010, 2011).
Zalameas transitory ontology
Zalamea insists on employing a terminology of different
forms of ontology (local, regional, transitory). Mathematics, between 1950 and 2000, as he adequately demonstrates,
proceeded by an analysis of streams, transits and deformations of structures, and their limits. A network was therefore
built, a web weaving together via passages and transits, but
also dualities and limits a bewildering variety of constructions. In such a web even Logic and Proof Theory find a new
structural significance,
where pivotal statements in logic such as the Loz theorem for ultraproducts, the completeness theorem for first-order logic, forcing
constructions in sets, and theorems of type omissions in fragments of
infinitary logic, can all be seen, uniformly, as constructions of generic
structures in appropriate sheaves. (284)

Indeed, sheaves constitute a structure of particular interest,


very often mentioned in the text. Born with Lerays analysis of
indexes and converings of differential equations, sheaves
238

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
are precisely what help to capture (and glue together) the continuous variation in the fibers. (285, n. 345). Moreover they
allow movement between the local and the global. So, thanks
to Grothendiecks generalization (sheaves on a Grothendieck
topology), they allow the integration of a profound web of
correlations in which aspects both analytic and synthetic,
both local and global, and both discrete and continuous are
all incorporated (286). Obviously, the category-theoretical
framework is the most fitting for this organization of mathematics. If in the Category of Sets objects are non-structured and
non-correlated conglomerates of elements, category theory
studies objects through their external, synthetic behavior, in
virtue of the objects relations with its environment (288).
Avoiding set-theoretic absolutes, in Category Theory the
notion of universality, for example, is relativized, becoming a unicity relative to given structures, in the given class
of morphisms. We have already observed how the analytic/
set-theoretic approach leads, perniciously, to the description
of every categorial diagram in terms of equations. Now the
constructions (co-product, adjuctions, pull-backs) or the
proofs in Category Theory can be based upon, and have a
meaning thanks to, symmetries and dualities present in the
diagrams, absolutely invisible in the equations. I therefore
once again underscore the fundamental contribution of
Noethers theorems, which extract physical invariants by
reading symmetries in the equations (of motion): in the
same way that categorial diagrams extract meaning out of
mathematical correlations, which then become visible and
comparable symmetries.11
11 We should note that the notions of scheme from algebraic geometry, of
frame of locale theory, or of Grothendieck topos, and their properties,
are not captured by an approach in terms of space = set + topology (or
space = set + structure). For example, from the constructivist point of
view, important theorems like Heine-Borels do not hold in set-theoretic
contexts, while they do in adequate, point-free, topos (see Cederquist and
Negri 1996). Similarly, constructions based on pull-back, insofar as they
are eminently categorial, allow to distinguish the obtained structure from
the set of points (when it is not an invariant with respect to the sets of
points in question). And a pull-back, typically, has a meaning a visible

239

Speculations VI
Zalameas work aims at moving the web of mathematical structures that have been introduced by contemporary
mathematics to the level of epistemological analysis, similarly as we saw the transfer the methodological content of
Einsteins invarintentheorie to a foundational approach.
That is to say, it aims at the construction of a comparative
epistemology, a sort of epistemological sheaf, sensitive to the
inevitable complementary dialectic of variety and unity that
contemporary mathematics demands (296). A mathematical knowledge some of whose highest peaks Zalamea (296)
enumerates (Grothendiecks motifs beneath the variations
of cohomologies [p. 144-148] [] Freyds classifying topoi
beneath the variations of relative categories [p. 245-246]),
proceeds between conceptual networks and their deformations
by means of series of iterations in correlative triadic realms:
differentiation-integration-invariance, eidos-quidditas-arkh,
abduction-induction-deduction, possibility-actuality-necessity,
locality-globality-mediation (297). The goal is that of a sort
of epistemological sheafification, where the local differential
multiplicity is recomposed into an integral global unity (299).
Is this a foundationalist epistemological analysis? It
surely is, in my opinion, since every epistemology is also
an analysis of a network of correlations and an history, a rational reconstruction of a constitutive path, evidencing the
network of passages and transits and, in this way, the unity of
the construction of knowledge. Of course, such an analysis
doesnt propose logical or ontological absolute foundations,
since the network is held together thanks to its own structure,
but also thanks to its friction upon the world, thanks to the
unity of language, thanks to its history through which it
constituted itself and thanks to the windows of intelligibility that it bestoys upon us. In this sense, to be provocative
once again, I would go as far as to say that mathematics helps
us to construct objectivity precisely because it is contingent,
the result of the history of a real friction with the world.
meaning only if we can appreciate its symmetries: the construction
itself is given by a duality (a symmetry) upon diagrams.

240

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
In this history we need to include that cognitive rooting, all
the way back to its pre-human form, at which I hinted before
when considering the role of memory in the constitution of
invariants. Zalamea briefly refers to another interesting and
technically deeper cognitive analysis, correlated to Gestalt,
with which Petitot (2008), and Citti and Sartri (2013) describe
the visual brain, neurogeometry. In the construction of the
world (in its friction with it) the brain, always active and plastic,
structures itself in a way that can be grasped geometrically,
thanks to complex symplectic structures. The brain organizes
the world through vision by imposing contours, correlating
points with the regularity of minimal forms, relative geodetics, and reading and imposing symmetries.
These kind of analyses, like those I mentioned above vis-vis memory, are not operations of cognitive reduction,
but rather tend to highlight the possible initial steps of a
constitutive path through which our communicating community has assembled conceptual mountains in a contingent, because historical, way. An alien friend of mine, from
the Sirius system, has no corporeal symmetry and interacts
with her ecosystem thanks to zuzrbs, and organizes her
world on the basis of a fundamental regularity that we cannot appreciate, but that may nevertheless be singled out, the
tzsuxu. It is another gaze, another epistemically efficacious
perspective, one perhaps compatible with ours (or even able
to unify microphysics and astrophysics, still, for us, objects of
incompatible descriptions). Another light is thus shone upon
the universe, of which we see little more than the humble tick,
whose Umwelt is so adroitly described by Von Uexkll (1934),
a tick who has been successfully coping with the universe for
far longer than we have.
Zalamea, instead, insists much on
the hypothesis of a continuity between the world of phenomena, the
world of mathematical (quasi-)objects associated with those phenomena, and the world of the knowledge of those objects which is to say,
the hypothesis of a continuity between the phenomenal, the ontic and
the epistemic From an epistemological point of view, the distinct

241

Speculations VI
perspectives are nothing other than breaks in continuity. (304-5)

I will leave it to the reader to adjudicate whether or not


it is possible to move with continuity between our two
points of view, and with mutual enrichment. As for myself,
I will insist, in the next section, on the critical transition
between these worlds, which needs to be analyzed in terms
of physical measure, or ways of access to phenomena. I have
indeed spoken of the constitution of invariants that lies at
the heart of the construction of (physico-mathematical)
knowledge, in continuity with action upon the world, yet not
with the world in itself.
I am in complete agreement with the project of a geometricization of epistemology [] that would help us to
overcome (or, at least, to complement) the logicization of
epistemology undertaken throughout the twentieth century (307). The distinction between principles of proof
and principles of (conceptual) construction (in Bailly and
Longo 2011) and the comparative analysis of the two sets
of principles in mathematics and physics first, and in biology, is precisely aimed at overcoming (complementing) the
monomaniacal (if profound and fertile) approach to Proof
Theory as the only locus for the foundations of mathematics.
And this geometry of epistemology consists, in primis, in
a Grothendieck-Lawvere-style geometrization of logic (but
one that also follows from Girard and his geometry of proof
[2001, 2007]). A project analogous to the geometrization of
physics, from Poincars geometry of dynamical systems
to the enormous work that goes from Riemann to Einstein
and Weyl in physics and from Gromov and Connes in quantum mechanics. We speak, therefore, of the construction of
mathematico-philosophico-metaphorical tools which, as
Chtelet puts it (paraphrased by Zalamea) in his historical
study of the nineteenth century,
in this search for a continuous articulation, include dialectical balances, diagrammatic cuts, screwdrivers, torsions, and articulating
incisions of the successive and the lateral, which is to say, an entire

242

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
series of gestures attentive to movement and which inaugurate dynasties of problems and correspond to a certain fluid electrodynamics of
knowing. (309)

Merleau-Ponty speaks of a glissment du savoir, in both


space and time: the epistemological challenge is to structure
and organize such knowledge, to give meaning to the moves
of both space and time in an historical and human sense of
knowledge, and consequently fostering the creation of new
perspectives, including new scientific perspectives.
To sum up, consider that in mathematics, in Zalameas words
the notion of sheaf, in a very subtle manner, combines the analytic
and the synthetic, the local and the global, the discrete and the continuous, the differential and the integral [p. 285-288]. In this way, the
sheafification of the analysis/synthesis polarity generates a new web
of epistemological perspectives. (319)

Zalamea presents his Platonism accordingly: not static,


but processual and methodological, so that the definitions
of mathematics, in reality, define methods; in no way do they
define existent things or simple properties inherent in such
things (330). This outlook mirrors my own stance on the
matter, and it is precisely that which allows us to pose the
problem to what extent such methods are to be preserved
and to what extent they are to be enriched or modified, when
moving to the interface between mathematics and biology
(Longo and Montvil 2014) and to what extent our attempts
of theoretical objectification of the living can still be inscribed
within this framework. The notion of mobility of the base
to which Zalamea refers, is close to the vision of objectivity
and effectiveness of mathematical construction upon which
I insist, insofar as it is the result of a phylogenesis and of a
human history: as the Platonic mobile base suggests, neither
invention nor discovery are absolute; they are always correlative to a given flow of information, be it formal, natural or
cultural (333). Which base changes should be operated in
order to move from the interface between mathematics and
243

Speculations VI
physics to that between mathematics and biology? From the
epistemological point of view, but also from that of an original
scientific construction, we are not interested in an ontology
of the transcendence of mathematical objects, but rather
in their transcendental constitution, as the phenomenologist would have it that is, their constituting through (and a
transit upon) the praxis of life and knowledge internal to
mathematics and often (an in a particularly fecund manner)
located in the interface with other forms of knowledge.
By posing the question of the relationship between mathematics and biology, therefore, I do not exclude a certain
autonomy of pure mathematics and of its effects on the
world. I want to stress, however, that mathematics has always
nourished itself on new interfaces, on new problems to which
new theoretical answers needed to be formulated. Thus, the
fluid electrodynamics of knowing can take us very far from
the original frictions, and an innovative metaphysics can
further fluidify this exchange just think of the role that the
philosophies of Nicola Cusano and Giordano Bruno, as well
as the practices of the painters of Italian perspective, played
in helping us to think the mathematical infinite and, in general, to conceive of new symbolic constructions of science
and mathematics (see Petitot 2004; Longo 2011b; Angelini
and Lupacchini 2013).
Regarding the relationship between culture, arts and mathematics, and their capacity to interact through the creation of
perspectives and points of view, Zalamea borrows Deleuzian
themes, and quotes at length an art historian, Francastel. On
these themes I want to remember Arasse, a disciple of Francastel and historian of painting, from whose more refined
analysis of the aesthetico-epistemological role of Italian
perspective I suggest we draw precious insights regarding the
play between the (local) detail and (global) sense of a painting, the interaction between painting and knowing artistic
subject (see Arasse 1999, 2009; S. Longo 2014), as well as the
sense of the (mathematical) infinite in renaissance painting.

244

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
The breath of aesthetics permeates mathematical creativity on at least
two levels, as detonator and as regulator. Referring to the artistic imagination, Valry writes in his Cahiers: Imagination (arbitrary construction)
is possible only if its not forced. Its true name is deformation of the
memory of sensation [] We have seen how contemporary mathematics
systematically studies deformations of the representations of concepts.
[] The visions of cohomologies everywhere in Grothendieck [p.
146], of groups everywhere in Zilber [p. 256], or metrics everywhere
in Gromov [p. 259], ultimately answer to a new aesthetic sensibility,
open to contemplating the local variations of (quasi-)objects through
global environments of information transformation. The aesthetic
regulation that allows the invasion of cohomologies, groups or metrics
be calibrated is decisive. (372-3)

Number and the Question of Measure


When three stones are lying on the ground and a volcano
spits out other two stones, neither the number 3, nor the
number 2, nor the concept of sum are there there are some
stones on the ground, and thats it. These will be five stones
for the practical action of whatever being decides to cut them
apart from their background, as we do (unlike the tick, for
example).
When a lion, in a group of three or four, hears five or six
distinct roars in the distance, it prudently changes course, in
order to avoid an uneven conflict or so the ethologists tell
us. The lion isolates an invariant of praxis, a praxis wherein
memory helps it to compare different active experiences,
from vision, hear and smell. However, the lion does not possess the concept of number, it merely builds but this itself
is no mean feat an invariant of action.
When we make the difficult, and very human, gesture of
an open hand with five outstretched fingers symbolizing a
numerical correspondence, and we refer to it in language, we
are giving ourselves the concept, furtther stabilized in writing.
Number is not already inscribed in the world, not even in
the discrete material of the stones on the ground, not before
245

Speculations VI
they are isolated from their background pragmatically as
many animals know how to, as well as in mythical-theoretical
manner, through language, as we have learned how to.
Number is not to be located in the biological rhythms that
regulate the time of the living either (Chaline 1999; Longo and
Montvil 2014). What is however interesting is the association
that Brouwer makes between the construction of the concept
of number and the two-ness of temporal discreteness: that
moment which passes by and becomes another (Brouwer 1975)
in the discrete succession of a musical rhythm, the rhythm of
the living, a proposal that evokes the Pythagorean intuition of
number and music. This picture is incomplete though: only a
plurality of active experiences permits the constitution of an
invariant, of that which does not change in the transformation
of one experience into another. The rhythm that organizes
time into the discrete, the small counting (the comparing
and counting of small quantities) which we share with many
animals (see Dehaene 1998), the spatial ordering of different
objects, together with the sense of movement associated with
order (Berthoz 1997) all of these precede and contribute to
the constitution of the (practical and conceptual) invariant,
being different active experiences. The passage, the transit,
the transformation of one into the other are necessary in
order to produce the invariant. All Pythagoreanism, holding number as intrinsic in the world, is misplaced: a brain,
embedded in its preferred ecosystem the body of a human,
historical and dialogical being is needed, along with a plurality of praxis from which to distill an invariant in memory
and then produce (in language) number, in order to stabilize
a concept resulting from a practical invariance with a long
evolutionary history.
Such constituted invariance comes into play even more
when it comes to analyzing processes and dynamics, where
one needs to remember that in physics and, a fortiori, in biology there is nothing but dynamics. We need then to measure
this or that observable pertinent of the selected process, a
theoretical proposal, also fixing a moment of measure, and
decide a beginning and an end of the process a far more
246

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
complex act than that of counting five stones. So measure
necessarily is, because of physical principles, an interval.
Thermic and gravitational fluctuations, as well as quantum
non-commutativity, do not allow us to associate a number
with their dynamics and with the pertinent observables,
but only approximations, changeable intervals. There is no
intrinsic number in no physical process: it is we, through the
difficult gesture of measurement, who associate numbers with
certain dynamics, as couples, extremes of rational intervals, as
concepts and as writing, constructed in language. And then,
with an eminently mathematical passage to the limit, one
which took 2.500 years to be achieved in relative completeness, we have proposed numbers without jumps nor gaps,
the Cantorian continuum, one of many possible continua
where the intervals of measure could converge.
The mediation or interface between mathematics and
the world requires the selection of a frame of reference and
measurement, the production of a number which is not in
the world but which must be extracted or proposed in order
to organize the world. In some cases a structure, a geometry,
can organize the world without numbers, so to speak. Thats
precisely what happened in the various facets of the geometrization of physics, of which I spoke above from Riemann
to Poincar and Einstein, from Weyl to Connes structures
that were somehow derived, as I said, from the problem of
measurement (ruler and compass, rigid body, Heisenbergs
non commutative algebras). This method can also be found,
for example, in the symplectic geometrization of the visual
cortex (see Petitot 2008; Citti and Sarti 2013). But like the
others, even this organizational proposal, a proposal of intelligibility that justifies the co-constitution of Gestalt with
and within the world, must then allow us to analyze fluxes, to
study functionalities and the dynamics of vision, analogously
to physical processes. And so geometry too requires numerical measure, with all the characteristics I mentioned, as does
every access to the structures of geometrized physics with
its difficulties and limits: classical, relativistic and quantum
(and in this case, biological).
247

Speculations VI
The flat (unidimensional) computationalists who see
algorithms and numerical calculi as coinciding with the world
should first reply to the provocative question I addressed to
the Pythagoreans, (see Longo and Paul 2010 for a formulation
of it) since they seem not to care about the issue of whether
the fundamental constants of physics are computable real
numbers. How unfortunate that Plancks h is not a whole
number, with G and c whole multiples of it! Is that God
playing tricks on us? And these constants (approximated
invariants of measure and theory) are present in all the significant equations, those that define the alleged computable
functions of physical processes. We also suggested to fix h
= 1, a legitimate move, modulo some transformation in the
metric of energy or time, but then the computationalists are
not able to compute G or c as exact real numbers, stuck, like
everyone else, in the interval of the new measure. If I were to
go out on a limb, I would bet that the fundamental constants
are random real numbers la Martin-Lf (see Calude 2002),
that is, strongly uncomputable real numbers, since they have
a Lebesgue measure of 1 (probability 1) in every interval of
the reals. It should be said that randomness, for real numbers, is a notion that has a meaning only to the infinite limit:
these incomputable reals are therefore an asymptotic jeu de
hazard, an infinitary dice game, available to God alone and
this capable of convincing even Einstein.
I defined the partisans of the computational world as
unidimensional, since the question of dimension is at the
heart of their flattening of knowledge. A first way of being in
the world and of constructing the intelligibility of the world
with other disciplines, indeed, is to appreciate its dimensionality, in the entire semantic richness of the word. To
begin with, it should be observed that everything changes,
in biology but also in physics, with the Cartesian dimension. From Poissons equations of heat, a standout case, to
all physical and biological processes, the spatial dimension
within which a process is analyzed is fundamental: its fixing
precedes every theoretical analysis it functions as its condition of possibility, we should say with Kant. In general, the
248

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
choice of a Category or of a Topos and their embedding in a
relative universe of Categories, with transits, functors, and
natural transformations to move from one to the other, is
a fundamental theoretical passage.
Consider the poverty, in speaking about the world, of a
Category, that of Sets, as an alleged ultimate universe of
fundamentals of intelligibility, where the set R of the reals is
isomorphic with Rn: the dimension being irrelevant for the
analysis. Or, even worse, the parody of a universe postulated
by the computationalists: the Category of discrete sets and
computable functions, where N is isomorphic with Nn. These
isomorphisms are essential to the theories in question: in the
first case they allow us to speak of cardinality, in the second
they allow the definition of Universal Machine, one of Turings great ideas, which led to the production of compilers
and operative systems of informatics. Personally, I have found
technical work in this latter Category, and its Types (see Rogers 1967; Barendregt 1984; Girard et al. 1989; Odifreddi 1989,
1999) very interesting, as explained for example in Longo
and Moggi 1984. The second Category is also well correlated
to the first one, once some algebra is added to it (see Longo
1983). Computability and Types, from Church to Girard, are
at the origin of and are still capable of giving mathematical sense to the extraordinary machines we have invented;
we need, however, to always try and offer correspondences
between their category and others of different nature (see
Asperti and Longo 1991).
Yet there are still those who want to analyze the Universe,
the brain, and the organism (the latter being codified by the
discrete structure of DNA) by remaining within N and its
finite, isomorphic powers. Now, the minimal structure one
needs to assume in order to correlate mathematics and the
world is a topological invariance, that of dimension. So, if
we consider, on R, the so-called natural topology, that of
intervals, the structure forbids the absurd isomorphisms
mentioned above: an isomorphism between two topologically
open sets of two different spaces forces the same dimension
of these spaces, which is then a topological invariant. This
249

Speculations VI
is a simple but beautiful correlation between topology and
physical measure, since natural topology derives from classical physical measure, an interval. This allows us to come back
to what I mentioned above about measure, and how such a
topological invariance has no meaning upon the discrete,
where the access is exact, absolute, and far from any form
of measure and access to physical and biological processes.
When we hope to ground the intelligibility of the world upon
one-dimensional, codifiable mathematical universes, as the
strings of bits that codify an image on a computer screen, we
break the symmetries that make the world intelligible (Longo
and Montvil 2014; 2014a).
Synthetically, one could say that that which is geometric,
and therefore a fortiori categorial, is sensitive to coding:
form, structure, the diagrammatic Gestalt, and organicity are
not invariants of coding, their entire sense is lost by coding,
as instead are information or digital computation, where
independence of coding is their mathematical strength. It
is therefore licit to claim that no physical process computes
(Longo 2009). In order to build one such process, the digital
computer, we had to invent the alphabet, modern logic from
Boole to Frege, Hilbertian formalisms, and Turings and Gdels
formidable codings. We thus individuated a new fundamental
invariant, the notion of computable function, independent
from the formal system. We had to inscribe these calculations,
codify them in a machine with discrete states, and make the
latter stable and insensitive to the codings and fluctuations I
mentioned above, forcing an electromagnetic dynamics into
the discrete, channeling it into an exact interface. So every
process in digital machines can be iterated in an identical
manner, via the implementation, on structures of discrete data,
of term-rewriting systems, i.e. systems of alphabetic writing
and rewriting, the most general form of computability (see
Bezem et al. 2003). This is a massive amount of science and
engineering, which includes the Lambda Calculus, with and
without types (see Barendregt 1984; Barendregt et al. 2013)
to which we gave, with many others, a geometrical significance in adequate Topos, bringing them back to bear upon
250

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
that geometrical organization I insisted upon, far from the
monomaniacal obsession with the computable discrete. This
has been a part of the network of constructed relations, the
synthetic movement of thought which lies at the heart of the
construction of mathematical knowledge, rich in concrete
and historical friction with the world.
To sum up, number and its structures are not already in
the world, and neither is it effective computing, which
is nothing but the formal transformation of the writing of
number: it is expressed in systems of re-writing, transformations of alphanumeric writing, upon which a machine
can operate. Phenomena, in physics in particular, are on the
other hand organized by us through non-arbitrary principles
of intelligibility, among which conservation and symmetry
principles that have dimensions and pose the problem of
access and measure. More precisely, I want to recall how the
conservation of energy and momentum (that are theoretical
symmetries) allow us to write the Hamiltonian, from which
to derive, for example, Newtons equations a specific case
but of great historical importance. From these, indeed, we
can proceed deducing the orbits with Keplerian properties.
This backward reading of history (starting with NoetherWeyls symmetries, and going back to Hamilton, Newton,
and Kepler) makes us appreciate the beauty and unity of this
strongly geometric construction of physico-mathematical
knowledge. This holds even if the planets and the Sun are not
identifiable with a material point mass, even if the phenomenal continuum is not made of Cantorian points (see Weyl
1987 on this topic) and thermic and gravitational fluctuations
make physical trajectories different from mathematical ones,
especially when there are two or more planets (Poincars
problem). The system, then, is chaotic and unpredictable in
modest astronomical time-frames (see Laskar 1994). And
the mathematics of negative results, as Poincar rebutted
to Hilbert, makes such phenomena intelligible. Only on a
computer screen does a trajectory made of pixels even the
chaotic one of a double pendulum follow exactly the path
dictated by the numerical solutions of an equation and can
iterate it exactly a physical nonsense. The symmetries of
251

Speculations VI
a computational model are different from those of the continuum, as we observed (see Longo and Montvil 2014a).
So the digital trajectory quickly diverges from that of the
mathematical continuum and from the real one. Moreover, restarted with the same digital approximation, on the
same number, it repeats itself again and again, identical to
itself, in secula seculorum, something that never happens in
physics and even less in biology, a science of radically nonreversible and non-iterable onto-phylogenetic trajectories,
cascades of changes of symmetry: a science of correlated
variations (Darwin).
Towards Biology: Problems and Conjectures
1- Variation, Continuum
I already talked at some length of the revolutionary role, in
contemporary mathematics, of sheaves and pre-sheaves. These
allow, in particular, for the construction of a new outlook on
variation, on the continuum and on the relation between local
and global. It is thus possible to break free of the dictatorship
of a continuum qua set of points and punctual variables
which do not make jumps nor sink into gaps a beautiful
construction we owe to Cantor and Dedekind, one of the
most profound constructions of mathematics, but very far
from the continuum of phenomena. Weyl (1987) has already
explained how absurd it is to consider such a mathematical
universe as congruent with the phenomenal continuum
the temporal continuum in particular, which is certainly
not made of points. It is meaningless, Weyl argued, to isolate
in a point a present moment that is not there anymore (as
Augustine would have it), even if he admits that, at the time,
he was inevitably subordinated to that exact construction of
mathematics. Today, we can do better, even though Cantors
and Dedekinds construction is still profoundly entrenched
into our mathematical imagery, and it is indeed the common sense of every school-educated person. Attempts (that
252

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
of Lawvere-Bell for example, see Bell 1998) to introduce the
Topos-theoretic vision into university educational programs
have had, for now, scarce success.
Perhaps the very general form of variation (or sheafification, as Zalamea puts it) on a continuum not composed of
points (and without enough points, as morphisms of the
terminal object upon the one in question) can fall back upon
the phenomena and help us make intelligible the continuous
variation considered in biology, just as complex numbers
imaginary objects of algebra have helped us to understand
microphysics. I said that variation is (one of) the fundamental
invariant(s) of biology, and that the mesh of biological and
ecosystemic relations channels this variation and forces a
permanent determination of the local by the global (and
vice versa), in a permanent critical transition which, for the
time being, resists a general and efficacious mathematization.
It is not obvious how to apply new instruments such as
Grothendiecks in a theoretical-biological field, and I personally
know of no successful attempt to do so. I have not seen, and I
do not know how to bring about, a passage from set-theoretical
punctuality to the actions of non-commutative monoids
in Grothendieck topoi (2234) as applied to a satisfactory
theory of organisms: it may be a job for a next generation.
The first obstacle, following our approach, is the genericity
of the physico-mathematical object and the specificity of its
trajectories. The objects and the transformations in and on the
Topos have the physico-mathematical character of genericity
and specificity: this is reversed in biology, as we said, with a
duality which represents a major conceptual challenge.
What type of categorial, technical, duality can reflect this
theoretical duality and produce a new outlook on biological
phenomena? I would be wary of shortcuts and of the arrogance
of anyone who would master such a beautiful mathematics:
the living is an extremely hard subject matter, a difficulty of a
different kind than the one faced by the beautiful mathematics we have discussed. We must first appreciate the richness
of the Theory of Evolution, the only great theory in biology,
as recounted by many great contemporary evolutionists to
253

Speculations VI
observe the complexity of the embryogenesis of a flys leg, or
the possible embryogenetic bifurcations of a zebra-fish in
order to fully understand why the competent and honest
experimental biologist is unable to give an answer to 80%
of the questions that the theorist poses to her when visiting
the lab. This is not the case in physics.
Perhaps another duality can be more easily grasped
through new structures. From Hamilton to Schrdinger we
have become used to understanding energy as an operator
(the Hamiltonian, the Lagrange transformation) and time
as a parameter. I hold that this approach, in biology, should
be inverted: here time is the fundamental operator, constitutive of the biological object by way of its phylogenetic and
ontogenetic history, while energy is nothing but a parameter,
as it indeed appears in scaling and allometric equations (see
Bailly and Longo 2009; Longo and Montvil 2014). If we clear
our mind of the classical schemes in which Hamiltons and
Schrdingers operators and Paulis controversial theorem,
which partially formalizes the distinct physical role of energy
and time, (see Galapon 2002) are given, we can perhaps
begin to see the whole in a new, dual way, as required by the
phenomenality of the living by its historicity, in this case.
Another theoretical path that needs a new outlook in terms
of continuity, density (as the rational numbers in the reals)
and of analysis of the local vs. the global is that of extended
criticality (see Bailly and Longo 2011, Longo and Montvil
2014). Critical Transition Theory, in physics, is an extremely
interesting discipline born within the fold of post-War
quantum mechanics yet further developed also in a classical
form for analyzing phase transitions through the application
of (quite a bit of) mathematics. The dominant framework,
obviously formalized on Cantor-style real numbers, describes
the transition as punctual, and this punctuality is essential
to the methods of Renormalization (see Binney et al. 1992;
Lagus and Lesne 2003). These deal with a cascade of models
which describe changes of scale and of pertinent objects, with
a change of symmetries (both breaking and construction of
new ones) at the punctual limit of the transition, where the
254

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
local appears imbricated with the global. The most familiar
examples are the formation of a crystal or of a snowflake, the
para-ferromagnetic phase transition, and Isings transition,
all mathematized as punctual transitions.
The criticality of the living, on the other hand, is extended:
it is always in a state of phase transition, in a permanent
reconstruction of its internal symmetries and in correlation with the environment (see Longo and Montvil 2014).
Indeed, in an organism every cellular reproduction has the
characteristics of a critical phase transition, for internal reconstruction and of the surrounding tissue. And within the
cell itself, molecular cascades pass through critical values
which can similarly be seen as phase transitions. The slight
modifications that always follow it are part of adaptive biology, including ageing (the increase in metabolic instability,
oxidative stress). An organism is somewhat like a snowflake
which reconstitutes itself in permanence, partially modifying
its symmetries, jointly to the correlations with the ecosystem.
In short, an organism is not merely a process, a dynamics, but is
always in an (extended) state of critical transition, permanently
reconstituting local and global symmetries. An interval of
criticality can give some idea, as I am trying to convey it, but
the density that would be necessary to describe it cannot be
the point by point density of a segment of Cantors line in
respect to every pertinent parameter or if it is, it is only
so in an inadequate manner. In any case, renormalization
methods cannot be applied, as such, to a classical interval of
criticality. A reasonable objective could be that of replacing
the Cantorian interval with the variation in/of a point-less
(pre)sheaf, thus giving a representation of density adequate
to renormalization, suitably extended.
2 - Measure
I have already discussed the crucial role and the theoretical
and experimental richness of measure in physics, the sole
form of access we have to the world (including perceptual
255

Speculations VI
measure), an interface between mathematics and phenomena. In biology the situation is even more complex. In the
first instance, a difference must be drawn between in vitro
and in vivo, a difference which has no meaning in physics.
Moreover, over the last few years we have seen the development of refined techniques of three dimensional cultures:
cells or tissue fragments from an organism are developed in
collagen suspensions from the same organism, giving rise to
matrixes or parts of tissues impossible to observe in traditional
and bidimensional Petri dishes. Thus both observation and
measure are profoundly changed, as if (but not quite as if) we
were somewhere in between the in vivo and the in vitro.
In any case, the duality I examined between generic and
specific, between biology and physics, radically changes the
meaning of a measure. The biological object is not an invariant
either of experience or of a theory, unlike the mathematical
and physical object. It is specific and historical and, to a greater
or lesser degree, individuated. Of course, the individuation
of a monocellular organism or of a single cell in a tissue is
minimal compared to that of a primate. And yet a cellular
culture is prepared, by biologists, with a full awareness of the
history of cells: cells from a given tissue are labelled, and the
descendents are distributed with the utmost care throughout
the world in order to reflect, collectively, on the iterability of
an experiment in reference to the history (i.e. the specificity)
of each cell or tissue. The same goes for lab rats, labelled and
traced along families as offspring of a same couple, so that they
will have a common, or at least known, phylogenetic history.
In an ongoing project, between laboratory experience and
theory, Mael Montvil is working on a theoretical analysis of
what he calls the controlled symmetrization of the biological
object factually practiced in laboratories, in order to deal with
its specificity and to make it as generic as possible. One of
the consequences of biological specificity is that the Gaussian
distribution of a measure does not have the same meaning
that it does in physics. For example in physics, in general,
deviations from or situations marginal to the Gaussian can
be seen as noise and decrease, relatively speaking, with the
256

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
increase of the total number of samples. In biology deviations are specific cases that can have great significance
for (cellular) differentiation and speciation, and increase as
the number of samples grows: enlarging the samples from
one population of cells, or rats (or of humans) to another
may radically change the response (to a therapy, say), a major
experimental and theoretical challenge. Only the control
(the normal cell, or rat, used as control), an unknown notion in physics, can help us understand the significance of a
variation, which is biological variability. And I want to insist
that variability, in biology, is not noise: it is at the origin of
diversity and therefore of the biological resilience of an individual, a population or a species and that this takes place
even in a population with a small number of individuals:
even in a population of a few thousands, individual diversity
contributes to evolutionary stability.
Which mathematical instruments should we use, or create,
starting with contemporary mathematics that is to say, going
beyond mere systems of (at best non-linear) equations, and
statistical methods invented at the end of the 1800s? When
Connes proposed non-commutative geometry he stood on the
shoulders of early 1900s giants. A highly refined theoretical
work then transferred the problem of quantum measure to
Heisenbergs matrix calculus, correlated with Weyls algebras
and Hilberts spatial continua, both used by Schrdinger for
his equation. As in relativity theory, or perhaps even more
so, the problem of measure had produced an imposing
theoretical edifice. This is certainly not the case in biology,
where practically no theory, as far as I know, accompanies or
guides extremely stringent experimental protocols, whose
originality and rigour are truly astounding for the theorist
who happens to visit the laboratory.
In short, I believe that it is necessary to first clarify what
to measure means before being able to imagine a process of
co-construction of mathematics and biology in a way vaguely
comparable to what took place between mathematics and physics in the last four centuries. The physicalist who denies the
existence of a properly biological problem, or the Pythagorean
257

Speculations VI
who claims that number is already there, should look
elsewhere. To associate a number with five stones, six roars
or five fingers, i.e. to build an invariant, is a long historical
process. To associate it with a physical or biological process
is a task which lies at the heart of experimental work, and
represents a major theoretical challenge, in biology even
more than in physics.
Conclusions on Zalameas Book
14.3.2. For mathematicians, logical axioms delimit a playground. But
which game are we going to play next?
7.4.1. Desire, and the resistance of the object, are what mathematicians
ordinarily use to distinguish mathematics from logic.
7.5.1. Grothendieck is rather like the Freud of epistemology.
(Lochak , 2015)

I hope I have managed to give the reader an appreciation


of how the immense shadow of Grothendieck dominates
Zalameas book. A French mathematician, the son of internationalist revolutionaries, migrating throughout all political
turmoils in Europe between the Russian revolution of 1905
and the Second World War, Grothendieck comes to France
when twelve years old, while the latter war was raging. He
first lived with his mother, and then in hiding. His life is as
original as his mathematics (see Lochak 2015). Without going
into the mostly dramatic details of the first, it is interesting
to note how Grothendieck is the only one of eleven French
winners of a Field Medal, who have had their university studies in France, to have neither studied nor taught at the ENS
in Paris, yet another touch of originality.
Following Grothendieck, Zalameas book gives priority
to the structures of mathematics, to their transformations
and deformations, and to the construction of meaningful
258

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
invariants. Taking this focus on structures, invariants and
transformations as the way to do philosophy of mathematics
the philosophical sheafication I mentioned above we move
away from set theoretical, logicist and formalist absolutes
(still grounded on the myth of the discrete and the finite
as absolutes) programmatically outside of the world.
We should however add that Grothendiecks work goes
beyond these speculations on symmetries, invariants, and
transformations. He had an exquisitely refined sense of the
purity of a mathematical definition. He was able to avoid,
arguably as no mathematician before him could, every contingency in the structures and proofs he proposed. All his
notions intrinsically encapsulate, so to speak, the maximal
invariance of a concept, to the extent that there is no need to
prove it, by identifying the adequate transformations: they
are intrinsic to the definitions12.
Grothendiecks approach unifies remote constructions in
mathematics, by proposing invariants which are surprisingly
shared by groups, topological spaces, manifolds of different
sorts (differential, geometric ), and by constructing, as
bridging notions, new mathematical structures. It is more
than a unification by generality, as the new objects proposed
have an autonomous, robust and profound mathematical
structuring. This allows to circulate in mathematics and
to propose and transfer common mathematical meaning
to apparently unrelated mathematical constructions. As
Grothendieck observed, sheaves on suitably changing sites
allow the circulation between continuous and discrete structures beyond the the founding aporia of mathematics, to
put it in Thoms terms.
As Zalameas book reminds us with regard to physics, yet
pushing beyond Zalameas arguments, it seems to me that the
12A typical example is the notion of tale topology. It is defined on a
category as a category, whose objects are morphisms on which schemes
act (as morphisms): the topology thus is given in a relational way, which
forces its right level of invariance. The notion of Topos as well is given
in a category-theoretic way: these are sheaves on sites (a small category
with a covering) .

259

Speculations VI
fundation of mathematics must take nourishment from the
dialogue with the theoretical foundations of other disciplines.
Not only in the dimension of historical analysis, but also in
the positive work of scientific creation, where epistemology
becomes entangled with the analysis of the construction of
knowledge. This construction is the result of a protensive
gesture which organizes the world, rich with desire for
(knowledge of) the real and constitutive of the mathematical
object through which it can be made intelligible; a real which
resists and channels mathematical invention, together with
its history. The analysis of this protensive gesture, and of its
historicity, is part of epistemological reflection, qua analysis
of a construction in fieri. The wandering of mathematical
work beyond any relation with the natural sciences is yet an
essential component of this construction, even more so if it
gives rise to new spaces for creation, new correlations and
abstract structures like Set or a new category of pre-sheaves.
The mistake is to take one of these creations and put it back,
as ultimate foundation, as a kind of Cantorian paradise outside
the world. In doing so, one loses the meaning of the whole
edifice, a network of relations of intelligibility, by absurdly
turning it upside down and making it stand on (perhaps
unidimensional) feet of clay. I am not here insisting on the
exigency of fundations as locus of certainty, but rather on
the necessity of the analysis of conceptual and cognitive roots,
of structures of sense as correlations, tracing their constitutive and historical path (broadly construed, as to include its
pre-human dimension). This project is far from pursuing
those unshakable certainties sought by Hilbert in a time
of great non-Euclidean uncertainties: on the contrary, there
is nothing more uncertain than the cognitive foundations of
mathematics as uncertain as any biological or pre-human
dynamics, as uncertain as a physical measure. However,
drawing upon a plurality of correlations of knowledge, an
historical epistemology of the interface between disciplines
construes them as mutually supportive, as epistemological
and epistemic webs: networks of meaning where the meaning of one helps us understand and constitute the other. An
260

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
epistemology, moreover, that helps us discern, in an original,
critical and ever-renewed way, the road to be built ahead,
which is what matters most.
Grothendiecks unifying methodology, within mathematics, based on the construction of new and often complex
deep structures, is also a remarkable example within the
foundational analysis and the practice of other disciplines:
reduction, say, rarely applies, while unification by new and
difficult theories marked the growth of science.13 Science is
not the progressive occupation of the real by known tools, in
a sort of fear of the novelty, but the difficult construction of
new theoretical frames, objects and structures for thought,
conceptual bridges or even enlightening dualities, such the
specificity of the biological vs. the genericity of the inert, with
its major consequences for a close analysis of measurement,
as suggested above.
My analysis has been inevitably superficial and incomplete;
even Zalameas large book is incomplete when it comes to the
richness of contemporary mathematical invention. Zalameas
style, informal and philosophical, may irritate some readers,
due to what could be considered as frequent flights of rhetorical fancy. Personally, I find it an extremely efficacious way to
express the enthusiasm that such mathematical abundance
deserves. As for rigour, when it comes to those fields in which
I can claim some technical competence (Types, Categories
and Topos, Girard, Lawvere) it all seemed to me to be
presented in a coherent and pertinent way, within the limits
imposed by the limited space dedicated to the numerous
themes transversally touched by Zalamea, who demonstrates
an outstanding breadth of knowledge.
I would like, finally, to commend the two associated publishing houses that published this volume: Urbanomic and
Sequence Press. In this as in other publications as for example
13 Newton unified Galilean falling apples and planetary movements, by
inventing brand new mathematics and theories. Similarly, Boltzmann
unified mechanics and thermodynamics at the asymptotic limit of the
ergodic hypothesis and the thermodynamic integral. Connes aims at
the unification of quantum and relativistic fields by a reinvention of
(differential) geometry

261

Speculations VI
the forthcoming English retranslation of Chtelets book (an
extremely hard work as Cavazzini, who recently translated
it into Italian, knows all too well) they certainly seem to
favour the creation of a critical space, by promoting originality, and offering an alternative to debates as well-established
as they are sclerotized in an oscillation between this or that
Scylla and Charybdis, even when the latter approach would
promise immediate success and, therefore, an high Impact
Factor a factor that is having a very negative impact on
science (Longo 2014).

Acknowledgements
I would like to thank Fabio Gironi, for his careful and patient translation, and Robin Mackay, who for a long time has
followed, with great acuity, my philosophical speculations.
Andrea Maffei and Olivia Caramello made a competent reading of, and many comments to, a draft of this essay, still highly
incomplete in comparison to the mathematics they master.
I would also like to thank Sara Negri, Sara Campanella and
Sara Franceschelli for their critical observations. Sara (Longo)
has helped me understand the movement from Francastel to
Arasse in the construction of a point of view in art theory,
so close to its construction in the sciences.

References
(Longos papers are downloadable from: http://www.di.ens.
fr/users/longo/)
Angelini A., Lupacchini R. (eds.) (2013). The Art of Science:
Exploring Symmetries between the Renaissance and Quantum
Physics. Dordrecht: Springer.
262

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
Arasse D. (1999). LAnnonciation Italienne. Une Histoire de Perspective. Paris: Hazan.
Arasse D. (2009) Le Sujet Dans le Tableau. Paris: Flammarion.
Asperti A., Longo G. (1991). Categories, Types and Structures.
M.I.T. Press.
Barendregt H. (1984). The Lambda-Calculus: its Syntax, its
Semantics. Amsterdam: North-Holland.
Barendregt H. P., Dekkers W., Statman R., (2013) Lambda Calculus with Types, Cambridge: CUP.
Bell J. (1998). A Primer in Infinitesimal Analysis. Cambridge: CUP.
Berthoz A. (1997). Le Sens du Mouvement. Paris: Odile Jacob.
Binney J., Dowrick N.J. , Fisher A.J. , Newman M.E.J. (1992)
The Theory of Critical Phenomena: An Introduction to the Renormalization Group. Oxford: OUP.
Bailly F., Longo G. (2009). Biological organization and antientropy. J. Biological Syst., 17(1): 6396.
Bailly F., Longo G. (2011). Mathematics and the natural sciences;
The Physical Singularity of Life. London: Imperial College Press.
Blass A. (1983). Cohomology detects failures of the axiom of
choice, Trans. Amer. Math. Soc. 279: 257-269,
Blass A., Di Nasso M., Forti M. (2012). Quasi-selective ultrafilters and asymptotic numerosity. Adv. Math. 231: 14621486.
Bezem M., J. W. Klop, R.. Roelde Vrijer (2003). Term Rewriting
Systems. Cambridge: CUP.
Bitbol M. (2000). LAveuglante Proximit du Rel. Paris: Flammarion.
Bitbol M., P. Kerzberg, J. Petitot (eds.) (2009). Constituting
Objectivity, Transcendental Approaches of Modern Physics.
Dordrencht: Springer.
Brouwer L. (1975). Consciousness, Philosophy and Mathematics, in Collected Works vol. 1 Amsterdam: North Holland.
Calude C. (2002). Information and Randomness. Dordrencht:
Springer.
Chaline J. (1999). Les Horloges du Vivant. Paris: Hachette.
Chtelet G. (1993). Les Enjeux du Mobile. Paris: Seuil.
Cederquist, J. and S. Negri (1996). A Constructive Proof of the
Heine-Borel Covering Theorem for Formal Reals In Types
for Proofs and Programs, Lecture Notes in Computer Science,
263

Speculations VI
Volume 1158, 6275.
Citti G., Sarti A. (2013). Models of the Visual Cortex in Lie Groups,
Advanced Courses in Mathematics. Dordrecht: Springer.
Corfield D. (2003). Towards a Philosophy of Real Mathematics.
Cambridge: CUP.
Danchin A. (2009) Information of the chassis and information
of the program in synthetic cells. Syst. Synth. Biol. 3:125134.
Dehaene S. (1998). The Number Sense. Oxford: OUP.
Diamanti-Kandarakis E., Bourguignon JP, Giudice LC,
Hauser R, Prins GS, Soto AM, Zoeller RT, Gore A.C. (2009).
Endocrine-disrupting chemicals: an Endocrine Society
scientific statement. Endocr. Rev., 30, 293-342.
Edelman G., Tononi G. (2000) A Universe of Consciousness.
How Matter Becomes Immagination. New York: Basic Books.
Elowitz M., Levine A. (2002). Stochastic Gene Expression in
a Single Cell. Science 297, 1183.
van Fraassen, B. (1993). Laws and Symmetry. Oxford: OUP.
Frege G. (1980). The Foundations of Arithmetic. Evanston: Northwestern University Press.
Galapon E. (2002). Paulis Theorem and Quantum Canonical
Pairs: The Consistency Of a Bounded, Self-Adjoint Time
Operator Canonically Conjugate to a Hamiltonian with
Non-empty Point Spectrum, Proc. R. Soc. London A, 458
(2002) 451-472.
Girard J.Y. (1971). Une extension de linterpretation de Gdel
a lanalyse, et son application a lelimination des coupures
dans lanalyse et la theorie des types. In 2nd Scandinavian
Logic Simposium, ed. by J.E. Festand, 6392. Amsterdam:
North-Holland.
Girard, J., Lafont, Y. & Taylor P. (1989). Proofs and Types. Cambridge: CUP.
Girard, J. (2001) Locus Solum. Mathematical Structures in
Computer Science. Computer Science Logic, 11(3): 323542.
Girard J. (2007). Le Point Aveugle. Paris: Hermann.
Gdel K.(1958) Ueber eine bicher noch nicht benuetze
Erweiterung des finiten Standpuntes. Dialectica, 12: 280-7.
Goldfarb W. (1986). Poincar Against the Logicists in Essays in the History and Philosophy of Mathematics, ed. by W.
264

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
Aspray and P.Kitcher. Minneapolis: Minnesota Studies in
the Philosophy of Science.
Harrington, L. & Simpson S. (eds) (1985). H. Friedmans Research
on the Foundations of Mathematics. Amsterdam: North-Holland.
Hilbert D. (1901). The Foundations of Geometry. Chicago: Open
Court.
Husserl E. (1970) The Origin of Geometry in The Crisis of
the European Sciences and Transcendental Phenomenology.
Evanston: Nortwestern University Press.
Johnstone P. (1982). Stone Spaces. Cambridge: CUP.
Kanamori A. (2003). The Higher Infinite: Large Cardinals in Set
Theory from Their Beginnings. Dordrecht: Springer.
Kosmann-Schwarzbach Y. (2004). Les thormes de Noether:
invariance et lois de conservation au XXe sicle. Les ditions
de lcole Polytechnique: Paris.
Lagus M., Lesne A (2003). Invariance dchelle. Belin: Paris.
Laskar J. (1994) Large scale chaos in the Solar System. Astron.
Astrophysics, 287, L9 L12.
Lochak P. (2015) Mathmatiques et finitude. Kim, Paris.
Longo G. (1983). Set-theoretical models of lambda-calculus:
Theories, expansions, isomorphisms. Annals of Pure and
Applied Logic, 24:153188.
Longo G. (1988). The Lambda-Calculus: connections to
higher type Recursion Theory, Proof-Theory, Category
Theory. A short (advanced) course on lambda-calculus and
its mathematics. Notes based on an invited lecture at the
Conference, Churchs Thesis after 50 years Zeiss (NL),
1986, in Annals Pure Appl. Logic, 40: 93133.
Longo G. (2009). Critique of Computational Reason in the
Natural Sciences, in Fundamental Concepts in Computer
Science, ed. by E. Gelenbe, J.-P. Kahane, Imperial College
Press, pp. 4370.
Longo G. (2010). Interfaces of Incompleteness. Downloadable paper (original in italian,in La Matematica, 4, Einaudi,
Torino, 2010).
Longo G. (2011) Reflections on (Concrete) Incompleteness
Philosophia Mathematica, 19(3): 255280.
Longo G. (2011b) Mathematical Infinity in prospettiva and
265

Speculations VI
the Spaces of Possibilities. Downloadable paper.
Longo G. (2012). Incomputability in Physics and Biology.
Mathematical Structures in Computer Science, 22(5): 880900.
Longo G. (2014). Science, Problem Solving and Bibliometrics. Invited Lecture, Academia Europaea Conference on
Use and Abuse of Bibliometrics, Stockholm, May 2013.
Proceedings, ed. by Wim Blockmans et al., Portland Press.
Longo G., Montvil M. (2014). Perspectives on Organisms: Biological Time, Symmetries and Singularities. Dordrecht: Springer.
Longo G., Montvil M. (2014)a. Models and Simulations: a
comparison by their Theoretical Symmetries. Forthcoming, downloadable.
Longo G., Montvil M., Sonnenschein C., Soto A. (2014). Biologys Theoretical Principles and Default State. Forthcoming.
Longo G., Moggi E. (1984) The Hereditary Partial Recursive
Functionals and Recursion Theory in Higher Types. Journal
of Symbolic Logic, 49(4): 13191332.
Longo G., Paul T. (2010). The Mathematics of Computing
between Logic and Physics. In Computability in Context:
Computation and Logic in the Real World, ed. by Cooper and
Sorbi. London: Imperial College Press/World Scientific.
Longo S. (2014) Voir et savoirs dans la thorie de lart de Daniel
Arasse. Ph.D Thesis, University of Paris I.
Mac Lane S., Moerdijk I. (1992). Sheaves in Geometry and Logic.
Dordrecht: Springer.
Makkai M, Reyes G. (1977). First-order categorical logic. LNM,
Springer, Berlin.
Mancosu P. (ed.) (2008). The Philosophy of Mathematical Practice.
Oxford: OUP.
Monod J. (1972). Chance and Necessity. London: Vintage Books.
Negri S., von Plato J. (2001). Structural Proof Theory. Cambridge:
CUP.
Noble D. (2006). The Music of Life. Biology Beyond the Genome.
Oxford: OUP.
Odifreddi P.G. (19891999). Classical Recursion Theory, vol. 12.
Amsterdam: North Holland.
Frdric Patras, (2001). La Pense Mathmatique Contemporaine.
Paris: PUF.
266

Giuseppe Longo Review Essay on


Synthetic Philosophy of Mathematics
Petitot J. (2004). Morphologie et Esthtique. Paris: Maisonneuve
et Larose.
Petitot J. (2008). Neurogomtrie de la Vision. Modles mathmatiques et Physiques des Architectures Fonctionnelles. Paris:
Les Editions de lEcole Polytechnique.
Poincar H. (1892). Les Mthodes Nouvelles de la Mcanique
Celeste. Paris: Gauthier-Villars.
Raj A., R. van Oudenaarden (2008). Stochastic Gene Expression and its Consequences. Cell, 135(2): 216226.
Riemann B. (1873) On the Hypothesis Which Lie at the Basis
of Geometry, English Trans. by W. Clifford, Nature, VIII
(183, 184): 1417, 36, 37.
Rogers H. (1967). Theory of Recursive Functions and Effective
Computability. New York: McGraw Hill.
Rudyak Yu. B. (2008). On Thom Spectra, Orientability, and Cobordism. Dordrecht: Springer.
Sonnenschein C., Soto A.M. (1999). The Society of Cells: Cancer
and Control of Cell Proliferation. Dordrecht: Springer.
Soto A., C. Sonnenschein. (2010). Environmental Causes of
Cancer: Endocrine Disruptors as Carcinogens. Nat. Rev.
Endocrinol., 6: 363370.
Uexkl J. (2010). A Foray Into the Worlds of Animals and Humans: With a Theory of Meaning. Minneapolis: University
of Minnesota Press.
Weinberg R. (2014) Coming Full CircleFrom Endless
Complexity to Simplicity and Back Again. Cell 157, March
27: 267-271.
Weyl H. (1932). The Open World: Three Lectures on the Metaphysical Implications of Science. New Haven: Yale University Press.
Weyl H. (1987). The Continuum: A Critical Examination of the
Foundation of Analysis (in German: 1918). Translated by
Stephen Pollard and Thomas Bole. Thomas Jefferson University Press.
Weyl H. (1949). Philosophy of Mathematics and of Natural Sciences. Princeton: Princeton University Press.
Weyl H. (1952). Symmetry. Princeton: Princeton University Press.

267